The purpose of this paper is to show by computation and graphic representation the dangers to which modern ships are subject from loss of stability through injury at the waterline.
In setting forth my views and deductions I will be justified in assuming what is already admitted, viz., that in all ships that are not completely armored at the waterline, there will be great damage at least from small shot, and, therefore, considerable inflow of water above the slopes of the protective deck. Argument would hardly avail against this assumption, for the adoption of the protective deck is confessedly based on the proposition that the projectiles shall perforate the outside of the ship, but shall not pierce the protective deck so as to endanger her vitals.
This paper is in no sense an argument against the protective deck; the claim that it protects buoyancy is accepted, but it is desired to call attention more directly to the fact that in using the protective deck to protect buoyancy, we are doing it at the possible expense of stability.
The alternative is one that may readily be accepted, but it is the purpose of this paper to show that it entails the necessity of understanding the conditions of stability of ships in the damaged condition; the necessity, in fact, of studying stability.
It is manifest that even some naval officers confuse the quality of steadiness in a ship with that of stiffness; yet a commanding officer who fails to discriminate between these two qualities is very likely to lose his ship in the event of waterline damage,—a claim which will become clear later. Such an officer, reasoning from the steadiness of his ship in a seaway and from the assumed value of coal as protection, would doubtless reserve the coal above the protective deck until forced to use it, ignoring the conditions of stability of his particular ship, whereas the aggregate risk might be much less under different conditions of coal stowage. All would depend on the conditions of stability for that particular ship.
It is claimed by the highest authorities that, for good behavior in a seaway and for steadiness as a gun-platform, a ship should have a metacentric height of from 3 to 3 ½ feet, and that this should even be exceeded in ships liable to waterline damage in action. Armored ships have, without exception, barring mistakes in their calculations, been constructed on this principle, and as an example we will quote from the official description of the U. S. armored cruiser Maine: "The vital parts are protected from shot and shell by a belt of armor, sufficient in length to insure stability even if the ends are riddled above the underwater protective decks, extending from the ends of the belt to the extremities of the vessel. Her metacentric height is 3.45 feet."
It is a well-known fact that underwater protective decks constitute a great danger to a ship's stability and safety in case of waterline damage, and this stands acknowledged by the fact that special provision is made in heavily armored ships for a sufficiently large metacentric height, so that any waterline damage above the underwater decks will not endanger their safety.
But if it is of vital importance that such provisions against damage to stability should be adopted in ships partially protected by heavy armor, with much more force do similar provisions apply to a class of ships which have no waterline protection whatever, and which are likely to be riddled from stem to stern by the projectiles of rapid-firing guns. It seems, therefore, that such ships ought to have a comparatively larger metacentric height, so as to be able to endure at least some waterline punishment; but, as this would produce increased rolling and a diminution of their steadiness as a gun-platform while in intact condition, an increase in their metacentric height over that of the armored ships would probably not be advisable, and other means should be devised to partially protect their stability.
But, on the other hand, no valid reason can be advanced why such totally unprotected ships should have a smaller metacentric height than those that are heavily armored.
In order to secure certain qualities for these ships, viz., speed, coal endurance, etc., they have not only been left entirely unprotected against waterline damage, but to attain the former the proportion of beam to length has been decreased, and the waterlines have been made so sharp that their metacentric height has not only been reduced to a minimum for intact condition, but cases have occurred where such ships, after having been completed, had to be provided with ballast in their double bottoms to enable them to go to sea.
Stability should never be subordinated to speed in a fighting ship, and if she has not sufficient natural stability due to properly proportioned dimensions and form, it is simply the fault of the designer.
Too much importance seems to be attached to coal protection. There is no doubt that coal, and especially patent fuel, will add somewhat to the resistance which is offered, but it is absurd to regard the ship's fuel, which is carried for the express purpose of being consumed for the ship's propulsion, as a great feature of her protection. Coal used is no longer protective, and since it is likely to be used from the bunkers above the protective deck, in fact oftentimes must be used, its protective character not only fluctuates, but in some cases vanishes altogether.
It appears doubtful whether the advocates of coal protection have given the subject that consideration which its importance demands; judging from the methods and appliances in use for trimming coal from the upper to the lower bunkers, they have not.
Fig. I represents the coaling arrangements of the protected-cruiser class. In time of peace this may work well; but in action when the sides of the ship are riddled it cannot be used. In damaged condition the water will not only enter the bunkers above the protective deck up to the height of the still-water level, but on account of accumulation due to rolling, etc., will rise above it; and as the lower edge of the sliding door (which is used for trimming the coal below) is several feet below the still-water level, it will be impossible to open the door without getting a continuous rush of water into the lower bunkers and fire-rooms. Leak-stoppers and mats will not avail to check the inflow of water, they cannot be adjusted under the fire of machine and rapid-firing guns, and if they could, would in turn be riddled and become useless. It has been argued that there is no necessity for trimming coal during action, as actions now will be short and decisive; but it may have to be trimmed below before the action commences, in which case what has become of its protective quality?
The necessity of chasing an enemy may arise after a ship has been riddled, and if the coal cannot be trimmed below—and it cannot with the present appliances—the chase has not only to be abandoned, but the ship is herself disabled for want of coal. Or, if chased in damaged condition, the ship would fall an easy prey to the enemy if her coal could not be trimmed below.
Conclusion: it is evident that coal cannot be made to serve the two purposes of protection and propulsion simultaneously, and is at best only a partial protection for stability. The apparent remedy is to slightly reduce the surplus coal supply, and to substitute for the weight and space so saved a better protection to stability in the form of properly constructed water-excluding belts. This being done, it may be possible to trim coal from the upper to the lower bunkers during action.
In considering the entrance of comparatively small quantities of water above the slopes of the protective deck, it is apparent that popular opinion regards the case only as one in which a small loss of buoyancy is the worst result; whereas the danger is likely to be critical, not from loss of buoyancy, but from the effects of the small quantity of water on the stability.
A most important question in the design of unarmored ships of war is that of protection to their stability; and a remedy lies in the application of suitable belts of light water-excluding material along the ship's sides.
Whatever this material may be—whether it be cellulose, woodite, or something not yet known or heard of—it is certain that the adoption of any such material is well worthy of consideration; and if a suitable material can be found which will even partially fulfill the conditions of a water-excluder, it will be certain to greatly prolong the life of an unarmored ship in action, and be of far greater value for ships with unprotected waterlines than the most minute subdivision of the waterline region into compartments, all of which may speedily be riddled by the small shot of rapid-firing guns.
Doubtless the next naval war will open our eyes to some important points, and upset many accepted theories in ship-designing, and it is to be presumed that this question of unprotected waterline in steel ships will be the most important and pressing one which the test of battle will bring before us.
Considering the unavoidable loss of stability of unarmored ships in action, their dimensions should be so proportioned as to give the largest possible metacentric height consistent with good behavior in a seaway and steadiness as a gun-platform, and it is certainly astonishing that in notable cases prudence has been violated, and that dimensions and other properties have been adopted which give a calculated metacentric height of only 2.10 or 2.20 feet, while in reality, through mistakes made in the calculations, it may fall short considerably. As it is the object of this paper to investigate the stability of "protected cruisers" in damaged condition, we are justified in adopting for the demonstrations a ship of foreign design, the metacentric height of which is officially given as 2.10 feet.
The official data for this ship, while meager, are sufficient for the purpose of the demonstrations for damaged condition.
From the above the metacentric height and other properties for the ship with full coal supply are easily calculated; the additional 450 tons of coal stowed on the slopes of the protective deck naturally raise the ship's center of gravity, and the metacentric height is reduced to about 1.72 feet, a loss of 4I inches. Fig. 3 represents the ship in the condition of full coal supply.
Again from this the metacentric height and other properties of the ship when the 400 tons of coal from the bunkers below the protective deck have been consumed, are easily ascertained. The consumption of this coal, with its center of gravity below that of the ship, must necessarily reduce the metacentric height, which is found to be 1. 1 2 feet only, a loss of .98 foot from that of the normal condition. This condition is represented by Fig. 4.
We will now proceed to demonstrate the effect of waterline damage for the three conditions of coal supply and stowage as shown in Figs. 2, 3 and 4.
In every case, except one, we will assume that the waterline damage is the same, and it will not be an exaggerated one if, in view of the fact that the ship is liable to be riddled from stem to stern, we assume that nine (9) compartments amidships, each of a length of twelve (12) feet, a total length of 108 feet, or less than one-third of the length of the ship, have been penetrated and laid open to the sea at or near the waterline.
We will first assume that the damage is confined to one side of the ship only, and that the fore and aft coal-bunker bulkhead has not been damaged in such a manner as to allow the water to pass through it.
Turning to Fig. 5, we find that the loss of buoyancy on account of water admitted into the bunkers on one side is equal to the area of the triangle abc multiplied by the length of the damaged compartments, which is 16 square feet X 108 feet = 1728 cubic feet or 49.37 tons, which would, if we imagine the ship to be held in upright position by some external force, occasion an increase in mean draught of 49.37/23.64 =2.10 inches, the tons per inch having changed from 25.70 to 23.64 on account of the loss of waterline area of 8 X 108 = 864 square feet. But there has also been a loss of moment of inertia. Had the entire waterline area for the length of 108 feet been lost, the loss in moment of inertia would have been y X t X 108, and as the mean half-breadth for the length of the 108 feet amidships is some what less than the greatest half-breadth of the ship, it can be taken as 23.70 feet approximately.
Conclusion: the fact that the comparatively small loss of buoyancy of about 50 tons, probably not m-ore than 1 ½ per cent of the reserve buoyancy, will capsize the ship, proves the danger of even comparatively small waterline damage associated with small metacentric height. With a metacentric height such as allowed for armored ships, the equilibrium would still have been stable, although accompanied by considerable heel.
Having shown that the penetration of the ship's side at or near the waterline for less than one-third of her length amidships makes the ship unstable, it is not necessary to consider the damage for both sides in detail; it suffices to state that if damaged on both sides for the length assumed, the ship would have a negative metacentric height of 2.72 feet, as shown in Fig. 6.
Conclusion: that even such a metacentric height as allowed armored ships would not save the ship from capsizing under the assumed damage and normal coal supply, and that the metacentric height should be assisted by special provisions for the protection of stability to insure the ship's safety even under the most moderate condition of gun-punishment.
Conclusion: that the admission of only 27 tons of water (about 1 per cent of the reserve buoyancy) on the slope of the protective deck will incline the ship more than twice the angle of depression of her guns, thereby making them useless on the intact side, whereas with an original metacentric height of 3.5 feet instead of 2.1 feet, the angle of heel would have been about 3° 35' 10" only, thus showing the great danger of too small a metacentric height for ships liable to waterline damage.
If, for the ship with full coal supply of 850 tons, we assume the same waterline damage of 108 feet on one side, the water admitted into the bunkers (when empty) would be equal to the area of the triangle.