The questions of the water-tight subdivision of ships and of safety of life at sea come before the public for a short period after each great wreck, and are then soon forgotten. They have excited more than usual interest in the last few years, owing to the great loss of life when the Titanic and the Empress of Ireland went down. The press and technical publications have given much space of late to considerations affecting safety of life at sea. Some of the suggestions brought forward are extremely foolish, while others are undoubtedly of much value and, if put into effect, could not but improve conditions.
The idea of subdividing a ship by water-tight bulkheads to give it greater safety in case of accident is an extremely old one. Admiral Belcher had water-tight bulkheads fitted in the British wooden sailing vessels, Etna and Terror, in 1830, which later prevented the latter vessel from foundering when her side aft was crushed in by ice. Her after bulkhead held, and enabled her to reach port in safety. The value of water-tight subdivision does not seem, however, to have been generally recognized until steamers came into use.
Mr. Williams, founder and manager of the City of Dublin Steam Packet Co., strongly recommended the use of water-tight bulkheads on iron vessels. In a paper on "Improvements in the Construction of Steam Vessels," read in 1837 before the Mechanical Section of the British Association for the Advancement of Science, he said: "When it is considered that those casualties which too often end in the sinking of a steamer are local in their origin, and affect but a small portion of the vessel, that the water admitted was often so small in extent as to be almost within the power of the pumps, it will at once suggest the importance and the efficiency of the protection by confining the water to that section of the vessel which has sustained the injury…. The plan is not restricted by any patent, and all are free to adopt it, and I expect hereafter to see this principle so adopted and improved that the security of steam vessels will keep pace with the greater utility of steam and extension to which they seem destined." Mr. Williams further stated that he had introduced water-tight bulkheads into many steam vessels, and, that he had practically and experimentally proved their efficiency.
The next great advance in the water-tight subdivision of Ships occurred when the Great Eastern was built in 1859. She was
{figure}
FIG. 1.-SECTION OF "BRITTANIA" BRIDGE
{figure}
FIG. 2.-MIDSIIIP SECTION "GREAT EASTERN"
designed by a French civil engineer, Brunel, with the aid and advice of Scott Russell. She was modeled, after the Brittania Bridge, as can be readily seen from Figs. 1. and 2. In the Great Eastern the inner bottom was more complete than on any merchant ships built until recently. She was much larger than any ship built before or for many years after her, having a length of 680 feet. Although she cannot be regarded as commercially successful, she has given to succeeding generations many valuable lessons in the construction of iron and steel vessels. Had her design ,been more closely followed in succeeding years, it is probable that the loss of life at sea would have been greatly reduced. In 1861, Mr. Charles Lungley read a valuable paper on "Un
{figure}
FIG. 3
{figure}
FIG. 4
{figure}
FIG. 5
sinkable Iron, Ships" before the Institution of Naval Architects. He objected to the fitting of numerous transverse bulkheads in a ship in the following words: "Such bulkheads also interfere very greatly with the stowage of the ship, and it often happens that the passages which are made through them, and are furnished with doors or valves, are found open at the very time when the safety of the ship depends upon the bulkheads being watertight.
Many serious disasters have occurred from the last cause." Mr. Lungley conceived the idea of so subdividing the ship as to avoid doors near or below the water line in the main water-tight bulkheads. He advocated the placing of a water-tight flat at about the level of the water line, extending throughout the length of the ship, except in engine and boiler compartments. In addition, several transverse and longitudinal water-tight bulkheads extended a safe distance above the water line. Access to compartments so formed is only through water-tight trunks extending above any probable water line at which the ship might float in case of damage. A sketch of the model with which he illustrated his paper is shown in Figs. 3-5. A is a water-tight deck, and B, B are trunks by which lower compartments are entered and through which the cargo is passed. In the discussion of the paper, objection was raised to the design on the ground that it would be very difficult to load and unload the cargo through the water-tight trunks. It is probably due to this defect that the design was not generally adopted. The importance of Mr. Lungley's paper lay primarily in his suggestion to abolish doors on a low level in the main transverse bulkheads. It was not until 43 years later that his suggestion was followed in the Dreadnought.
In spite of the experience gained and of the advice of many eminent naval architects and ship-builders, the requirements of law relative to water-tight subdivision seem to have been regrettably lax. In 1854 the English Board of Trade required that “every steamship of over too tons be divided by substantial transverse water-tight partitions, so that the fore part 9f the ship shall be separated from the engine-room by one of such partitions, and so that the after part shall be separated from the engine-room by another such partition." This act remained in force until 1862, when it was rescinded. After that date the Board of Trade required no subdivision whatever, and would issue a certificate to a ship consisting merely of a hollow shell.
The committee of Lloyd's Register of Shipping had much more stringent. regulations. As Lloyd's registered, then as now, the great majority of the world's shipping, the action of 1862 of the Board of Trade did not have such it deleterious effect as might have been expected. The requirements of Lloyd's Register did not, however, insure as great a degree of safety as might reasonably be required. In 1876, ship-owners in England were asked to send to the Admiralty the names of any of their ships which were so divided that they would not necessarily sink if bilged in their largest compartment, when loaded to their normal load water line with cargo in their hold. The number of vessels found so qualified to be placed upon the Admiralty list of merchant naval reserve, was only 30. Conditions improved somewhat in the next few years, so that by 1883 there were 300 ships on the Admiralty list. The requirements of Lloyd's for vessels over 300 feet in length would in general comply with the requirements of the Admiralty.
In 1890 the Board of Trade appointed a committee of experts to investigate the subdivision of ships and strength of bulkheads. This so-called Bulkhead Committee advised the adoption of requirements much more severe than were then in force by Lloyd's.
The recommended requirements were such as to insure the safety of passenger ships longer than 425 feet, if any two compartments were flooded. The committee also drew up a set of tables to be followed in the construction of bulkheads.
The recommendations of the Bulkhead Committee caused much unfavorable comment, the concensus of opinion at the time being that they were much too severe, especially those relating to the construction of bulkheads for strength. The recommendations were not enforced, and have sine been adopted on very few ships. Recent experience seems to indicate that they were not severe enough, and that bulkheads constructed according to the tables of the committee have not sufficient strength.
Up to this time, 1890, little attention had been paid to the question of water-tight doors in bulkheads, although many ships had been lost, directly or indirectly, through failure of such doors to be closed at critical films.
Loss OF THE " VICTORIA "
On June 22, 7893, during maneuvers of a British squadron in the Mediterranean, H. M. S. Victoria was rammed and sunk by the Camperdown. The squadron was in line of divisions with a distance of about 1200 yards between the divisions, when the signal was made from the Victoria: " Second Division alter course in succession 16 points to starboard, preserving the order of the fleet. First Division alter course in succession 16 points to port, preserving the order of the fleet."
The distance between divisions was too short for such a maneuver to be made, and the two leading ships came into collision, the Camperdown striking the Victoria on the starboard bow at an important transverse bulkhead and at an angle of about To degrees abaft the beam. At the time, the Victoria was making a speed of about 5 knots, and the Camperdown, a speed of about 6 knots. The protective deck of the Victoria locked the stem of the Camperdown and prevented any ripping action which might have resulted from the speed of the former. The spur of the Camperdown was driven about 9 feet within the side plating of the Victoria at a depth of about 12 feet below the water, destroying the watertightness of the transverse bulkhead, at this place. The Camperdown swung aft through an angle of 20 degrees and backed out clear about one minute after she struck. The Victoria settled gradually by the head and took a list to starboard for nine or ten minutes, when water began to flow down through the open gun ports in the turret and through. the 6-inch gun ports and the forward door of the casement. This quickly destroyed the stability by reducing the moment of inertia of the plane of flotation. The ship took a sudden lurch to starboard and capsized. Immediately before the lurch took place the upper deck at the bow was 13 feet under water, having been depressed about 23 feet.
One minute before the collision occurred, the captain of the Victoria gave orders to close water-tight doors. This was not sufficient time for the execution of the orders, and the men were driven back by the inflow of water immediately after the collision, leaving many doors in the forward part of the ship open or not properly closed. The doors had just been overhauled and were in good condition; the damage due to the collision did not seem to make any of them inoperative.
{figure}
FIG. 6.—H. M. S. " VICTORIA "
After extensive investigations, calculations and experiments, the Admiralty came to the conclusion that had all water-tight doors and hatches been properly closed the ship would have changed trim only 13 feet and have taken a heel of about 9 degrees to starboard. The ship would have retained ample transverse stability, the metacentric height being 2f feet. The upper deck at the bow would have remained above water and the ship would have been manageable.
In Fig. 6 are shown the water line if the water-tight doors and hatches had been closed, and the actual water line just before the turret ports were flooded. If these ports, the 6-inch gun ports and the doors in the forward end of the casement had been closed, the ship would not have capsized, and might have been able to reach shore, only 5 miles distant. At any rate she would have stayed afloat much longer than she did.
The loss of the Victoria cannot be attributed to the inefficiency of her water-tight subdivision. The forward part of the ship was minutely subdivided by two water-tight decks, transverse bulkheads, and a few side longitudinal bulkheads. Instead, we must charge the loss entirely to the failure to close water-tight doors, ports and hatches. Had there been no openings in the main transverse bulkheads, below the water line, the ship would undoubtedly have remained afloat.
Loss OF THE "REPUBLIC
On January 23, 1909, the White Star liner Republic, of 15,400 tons, built in 1903, was rammed by the Florida, a much smaller vessel. The Republic was struck on the port side, and a hole was torn above and below the water line abreast the engine-room. The engine-room was but 5o feet long, and was the only compartment flooded as a direct result of the collision. The passengers were removed from the Republic, and she was taken in tow. She settled gradually by the stern, showing that the engine-room bulkhead was leaking. Forty and one-half hours after the collision, the Republic's bulkheads gave way and she sank. The bulkheads, although only single-riveted, were constructed according to the best practice of the time when she was built. The loss of the ship clearly shows that bulkheads as constructed at that time were of insufficient strength.
LOSS OF THE " TITANIC "
The Titanic was a triple-screw liner of 46,300 tons gross, belonging to the White Star Line. She had 15 transverse watertight bulkheads, as shown in Figs. 7-8. Their water-tightness extended up to D deck abaft bulkhead K and forward of bulkhead B. Bulkhead A extended to C deck, but was water-tight
{figure}
FIG. 7-“TITANIC”
{figure}
FIG. 8-“TITANIC” - INWARD PROFILE.
only to D deck. The heights of D and E decks above the 34 ft. 7 in. water line are given by the following table:
HEIGHT ABOVE 34 FT. 7 IN. WATER LINE
Lowest part amidships At bow At stern
D Deck 20-0 33-0 25-0
E Deck 11-0 24-0 16-0
G deck ill the after peak tank, and the orlop deck in the fore peak tank and abaft the turbine engine-room (bulkhead M) were watertight. Elsewhere no decks were water-tight except over small areas. There was' a structure some distance in from the ship's side in the electric, machinery compartment forming six watertight compartments, used as fresh-water tanks. There were in all 29 water-tight compartments above the inner bottom.
There was an inner bottom about 5 feet from the outer skin, extending from bulkhead A to 20 feet forward of bulkhead P. For about half the length of the ship the inner bottom extended to a height of 7 feet above the keel, but at the ends of the ship it did not extend to such a height. The space between inner and outer bottoms was divided into 44 water-tight compartments. Bulkheads A, B and P were without openings. All the other bulkheads had water-tight doors. Bulkheads 0 to 0 inclusive each had a vertical, sliding, water-tight door at the level of the floor of the engine- and boiler-rooms.
The Titanic collided with an iceberg at 11.4.0 p. m. on April 14, 1912. The starboard side was ripped open at a height of 10 feet above the keel and for a distance of 300 feet. The fore peak, No. 1 hold, No. 2 hold, No. 3 hold and the forward boiler-room (No. 6) were damaged to such an extent that the inflow of water was beyond the capacity of the pumps. The second boiler-room (No. 5) was damaged so that water poured in as it would from an ordinary fire hose. The upper part of the fore peak was not flooded until the bow of the ship was submerged to deck C, when water poured down the scuttle, filling the compartment.
Since the ship was designed to keep afloat with two Of her main transverse compartments flooded, it is, easy to see that the injury received was a fatal one. The flooding of the five forward compartments caused the water to rise above the top of bulkhead E and to pour down into boiler-room No., 5 and fill it up until the water rose above the top of bulkhead F and flooded the next boiler-room, and so on until the ship sank. Tf the forward five compartments had been completely flooded,: the ship could still have been saved had the bulkhead deck (deck, E, amidships) been made effectively water-tight, as it had water-tight trunks, extending up to C deck, around all openings in the bulkhead deck.
No evidence was brought forward in the investigation to, prove that the water-tight doors did not function properly or had any influence on the sinking of the ship, or that any bulkheads failed. There is some doubt, however, as to whether the bulkheads, which were only single riveted, could have stood up long enough to enable the ship to reach port in safety under her own steam.
CONVENTION FOR SAFETY OF LIFE AT SEA
Following the Titanic disaster, an international Convention for the Safety of Life at Sea was held. It was signed by representatives of all the great powers in January, 1914. The articles are applicable only to merchant ships, mechanically propelled, carrying More than 12 passengers, and engaged in international trade.
Certain terms are defined by the convention as follows:
The bulkhead deck is the uppermost continuous deck to which. all transverse water-tight bulkheads are carried.
The margin line is a line drawn parallel to the bulkhead deck at side and 3 inches below the upper surface of that deck.
The permeability of a space is the percentage of that space which can be occupied by 'water.
The floodable length for a given point in a Ship, with a continuous bulkhead deck, is the maximum percentage of the length of the ship, having its center at the point which can be flooded (under certain assumptions) without the ship being submerged beyond, the margin line.
By the convention, definite permeabilities are assigned for the different compartments of a ship. The maximum permissible length of a compartment, having its center at any given point in the ship's length, is obtained from the floodable length by multiplying that length by an appropriate factor, called the factor of subdivision. In calculating the floodable length, the specified permeability for, that part of the ship is used. The factor of subdivision is a function of the length of the ship and the service on which she is engaged, and is taken from curves given by, the convention. The degree of safety required is, therefore very properly greater for large ships than for small ones and greater for passenger ships than for mixed passenger and cargo ships.
The convention requires fore and after peak bulkheads at appropriate distances from the ends of the ship,: It also requires that water-tight bulkheads shall ,be constructed so as to be capable of supporting the pressure due to a head of water to the margin line, but does not lay down any fixed rules or tables for the construction of such bulkheads.
The convention recognizes the desirability of cutting down the number of watertight doors to a minimum. It allows no doors below the margin line in the collision bulkhead or in bulk heads between cargo spaces. It does, however, permit one door in each main transverse bulkhead in the machinery spaces (including boiler and engine rooms).
In regard to inner bottoms, certain regulations are laid down, as follows:
Ships of 200-249 ft.: Timer bottom from machinery space to fore peak.
Ships of 249-300 ft.: Inner bottom from machinery space to fore and after peaks.
Ships over 300 ft.: Inner bottom from fore peak to after peak, extending out far enough to protect bilges.
Ships over 699 ft.: Inner bottom for at least half ship's length amidships and forward to the fore peak shall extend up ship's side to a height above the top of keel not less than 10 per cent of ship's moulded breadth.
It will be noted that the convention does not require any deck to be water-tight.
REPORT OF BULKHEAD COMMITTEE
At about the same time that the International Convention met, the Board of Trade appointed a Bulkhead Committee to investigate the question of water-tight subdivision of merchant ships.
This committee has investigated the subject in a very thorough manner, and in January, 1915, made public that portion of its report dealing with ships carrying more than 12 passengers. As
the convention received a preliminary report from the Bulkhead Committee and adopted many of its suggestions, it is natural that the recommendations as set forth in the committee's report should agree with those of the convention. There are, however, certain important differences.
As a result of tests on a cofferdam, the committee gained much useful information regarding the strength of bulkheads. Tables of scantlings, of brackets; and of stiffeners to be adopted in new construction are given, which are considerably in excess of the existing general practice.
The requirements as to the spacing Of transverse bulkheads are slightly more stringent than those of the convention. Distance between bulkheads is determined by a series of curves, making individual bilging calculations unnecessary.
The committee recommended the mean permeability of the whole vessel as a suitable criterion of service on which to base the factor of subdivision.
If a water-tight deck is fitted, it should not be below the level of the load water line at any part of its length. The committee recognizes the value Of such a deck against the effects of a long ripping blow, but does not consider it of great value against effects of collision. No increase in the spacing of transverse bulkheads is permitted when such a deck is fitted.
Longitudinal subdivision and inner skins extending up the sides are entirely optional even in the largest vessels. If they be fitted, it must be proved by calculation that the margin line will not be brought under water when any space bounded by a transverse water-tight partition between outer and inner skin or longitudinal 'bulkhead on one side is flooded for a length equal to the floodable length of the vessel in that region. The committee recommends, however, that the tank top plating be carried out to the ship's side at a height not less than that at the center, as specified by the classification society or other authority responsible for the scantlings.
CONCLUSIONS
In planning the water-tight subdivision of a ship we must have clearly in mind three factors, namely: Safety, profit and speed: Perhaps unconsciously, the owner is likely to place profit first and speed second, and to give only such consideration to safety as is required by law or as is necessary to keep down the insurance rate to a reasonable figure. Curiously, we cannot rely on the public to insist on safety above all else. Except for a short period after each great wreck, the public pays little attention to safety, and less to profit, but demands speed at all costs. Obviously a ship must be so designed as to make her operation profitable; otherwise she could not be run at all. In considering the question of safety, we must try to give the maximum safety which will still permit her to yield a reasonable profit. This maximum safety should clearly be a function of the size and service of the ship. Fortunately this principle may be worked out in practice for adequate subdivision is given to a large ship much more easily than to a small one, while keeping expense of building and operation down to a reasonable figure. It is also easy to give greater safety to a passenger ship than to a cargo ship. Speed of passenger ships should be regarded as a luxury to be paid for by those demanding it.
Bearing the above principles in mind, we should next consider the casualties against which the ship should be protected and the means we have at our disposal for accomplishing this. In general, the accidents endangering the life of a ship for which we should provide to a greater or less degree by watertight subdivision, are as follows:
(1) Being rammed by another ship.
(2) Ramming another ship.
(3) Grounding or running upon the rocks:
(4) Striking a derelict, iceberg, or other floating body.
(5) Flooding due to leaks in the skin of the ship.
(6) Flooding due to seas coming over the ship or through openings in the sides.
Of these the first two, because of the greater chance of their occurring, are by far the most important. They should therefore receive more consideration in the designing of the watertight subdivision than any of the other possible accidents. Grounding is much more likely to occur on ships engaged in one kind of service than on those engaged in another. For instance, the chance of a trans-atlantic liner grounding is very remote, while the chance of a tramp steamer running upon an uncharted rock is not small. The danger of a ship's striking a derelict, iceberg, or other floating body is not great. The author could find record of only three ships having collided with icebergs in the last 30 years. The flooding by seas coming through openings in the weather deck or in the sides of the ship is much more likely to occur in a small ship than in a large one, while the springing of leaks is caused principally by lack of structural strength in the hull of the, ship.
To give protection against the effects of these various accidents, we have water-tight bulkheads, inner bottoms, water-tight decks, or flats.
Water-tight Bulkheads.—There seems to be no question that all ships should be fitted with as many main transverse bulkheads as practicable. The distance between such bulkheads depends upon the size and service of the ship. In any case, such bulkheads should be carried to such a height above the water line that the upper edge will not be reached by the water in any case of flooding, after which the ship may reasonably be expected to remain afloat. These main transverse bulkheads protect principally against- the effects of being rammed. Since ramming is likely to occur at one of these bulkheads, they should, if possible, be close enough together to permit two main compartments to be flooded without the tops of the bulkheads going under water. Of course, for large passenger ships the spacing should be much closer. There should be one or two collision bulkheads forward, carried up one. or two deck-heights higher than the other main transverse bulkheads, and there should also be an after peak bulkhead. The object of these bulkheads at the. ends Of the ship is primarily to guard against the effects of ramming another ship.
The question of the advisability of having longitudinal watertight bulkheads is one upon which there is much difference of opinion. Longitudinal tenter-line bulkheads in the large compartments are to be avoided, since in cases of flooding they are likely to cause the ship to take a dangerous heel. Side coalbunker bulkheads are not generally open to this objection, and may prove very valuable in case of a long ripping blow, such as would be given by an iceberg or a derelict. In cases of collision between two ships; there is also sometimes this tearing action. Such was the case when the Grosser Kurffirst was rammed and sunk by the Klinig Wilhelm, the headway of the former causing considerable tearing of the bottom plating. Longitudinal bulkheads in the ends :of the ship serve to divide the space more minutely, and undoubtedly make for greater safety. There seems to be no objection to dividing the engine-room space into three parts by longitudinal bulkheads, as is done on the Aquitania. If fitted, longitudinal bulkheads should be carried water-tight either to some water-tight deck or to the same height as the main transverse bulkheads.
Inner Bottoms.--Nearly all ships have an inner bottom throughout the machinery spaces but extending only over the flat part of the bottom. Especially in long ships, the structure between the two bottoms is essential to get sufficient strength for the hull. The space so found is generally subdivided and used as ballast, oil, or reverse feed storage tanks. The depth of such compartments is ample for protection against the effects of grounding under them. With such an inner bottom, a ship is open to a great amount of flooding from a long ripping blow dealt anywhere from the turn of the bilge up to the water line. It therefore seems advisable in large liners to extend the inner bottom up the sides as far as the water line from the collision bulkhead to the after engine-room bulkhead. In places where there are side bunker bulkheads, this may not be necessary, but it still seems advisable for the largest ships. Side compartments in the double bottom may be of use in bringing the ship on an even keel in cases of dangerous heeling.
Another way of protecting the side of the ship from the turn of the bilge up is to fit wing bulkheads entirely independent of the ship's side framing, at a distance of about 8 feet from the side. The space outside the wing bulkheads would be regarded as double-bottom space, and could be entered only through watertight man-holes. Such an arrangement would of course be somewhat wasteful of space, but on large liners having transverse bunkers these bulkheads could without great difficulty be fitted throughout the boiler-rooms.
Water-tight Decks or Flats.—Water-tight decks or flats at or below the water line may be useful in confining the extent of the flooding in the case of a long ripping blow beneath them, or they may serve to prevent the tearing action in the case of collision. The protective deck of the Victoria was useful in the latter capacity. The bow has a water-tight deck at about the level of the water line which strengthens it and lessens the danger in case of ramming another ship. It is, however, the general opinion that water-tight decks are more useful when placed above water. On large liners it seems advisable to make the bulkhead deck effectively water-tight throughout its length. In order to do this, openings in the deck must either be fitted with water-tight hatches, or trunks must be fitted around the openings and carried to one or two deck-heights higher. This prevents the flooding of any main water-tight compartment, should a heel or a trim cause the bulkhead deck to be submerged.
Strength of Bulkheads.—It is of vital importance to construct all main water-tight bulkheads in such a manner that even in the most serious cases of flooding they will remain tight in fact as well as in name. This matter has not received sufficient attention until quite recently. It is to be hoped that the recommendations of the Bulkhead Committee will work a needed reform in this regard.
Water-tight Doors.—Water-tight doors in a main bulkhead are always a menace. Even if they are of the most approved type and are fitted with all imaginable safety devices, it is never sure that they will work when most needed. They may be warped by the effects of an accident or may be temporarily out of order, or may be blocked by something solid washed under them. These and many other things might happen to prevent them from closing. Their use below the bulkhead deck should be avoided if possible. In any case they should not be permitted below the water line in the main transverse bulkheads. Not to have doors between boiler-rooms is no doubt an inconvenience, but it is better to sacrifice convenience than to sacrifice safety.
EXAMPLES
The latest liners follow, in a general way, the principles laid down above. The Aquitania, the latest Cunard liner, Fig. 9, has side coal-bunker bulkheads together with inner bottom carried to the turn of the bilge. There is a water-tight deck carried at about the water line for the whole length of the ship. Forward and aft there are water-tight trunks for each cargo hold from this deck to the weather deck. The water-tight subdivision of this ship is almost exactly like that suggested by Mr. Lungley in 1861, Figs. 3-5.
Another system of subdivision is illustrated by the White Star liner, Brittanic, of 50,000 tons gross. She has transverse bunkers, and has an inner skin 30 inches from the outer shell, extending up to 6 feet 6 inches above the load water line throughout the machinery spaces. In addition to this she has 16 transverse water-tight bulkheads, of which five extend to the level of the bridge deck, 40 feet above the load water line.
The construction of these two liners illustrates the tendency on the part of ship-builders and owners to profit by experience, and to pay more attention to the proper water-tight subdivision of ships. The terrible lessons of the last few years have also gone far to dissipate the popular delusion that ships can ever be made "unsinkable."