IMS' 3 1924 105 726 123 In compliance with current Copyright law, Cornell University Library produced this replacement volume on paper that meets the ANSI Standard Z39.48-1992 to replace the irreparably deteriorated original. 2007 'Ml Cornell University Library The original of tiiis book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924105726123 BOUGHT WITH THE INCOME FROM THE SAGE ENDOWMENT FUND THE GIFT OF 1891 A^^.'bS.o ./.0/k.^.hi.., A TEXT BOOK NAVAL A'RCHITECTFRE FOB THE USE OP OFFICERS OF THE KOYAL NAVY. By J. J. Welch, Assistant Constructor, Royal Navy ; Instruotor in Naval Architecture at the Royal Naval College, Greenwich ; Memhcr of the Institution of Naval Architects. Published by order of the Lords Commissioners of the Admiralty. LONDON: FEINTED FOR HEE MAJESTY'S STATIONEEY OFFICE, BY DAELING & SON, LTD., 1-3, GREAT ST. THOMAS APOSTLE, E.G. And to be ptirohased, either directly or through any Bookseller, from ETEB AND SPOTTISWlODB, EAST HARDING Street, FLEET Street, E.G. : JOHN MENZIES & Co., 12, HANOVER STREET, EDINBUKGH ; and 88 & 90, WEST NILE Street, Glasgow ; or HODGES, FIGGIS & CO., 104, GRAFTON STREET, Dublin. 1891. Price Four SliMlingt. CORNELL UNIVERSITYI LIBRARY CONTENTS. Ohap. I'a<5'=- I. Buoyancy of Ships - .... 5 II. Stability of Ships - - - 1-' III. Oscillations of Ships - - 50 IV. Materials for Shipbuilding and modes of Connecting - . - OS V. Classification of Ships - 78 VI. Keels and Framing - - ^1 Til. Stems and Stbrnposts - - 'J8 Till. Bottom and side Plating and Planking - 102 IX. Deck Plating and Planking - - 110 X. Watertight Sub-division of Ships- - - 113 XI. Pumping, Flooding, and Drainage Arrange- ments - . - - - 121 XII. Ventilation of Ships - - 131 XIII. Steering Arrangements - - - 130 XIV. Protection of Ships against Gun Attack - 141 XV. The Preservation of Ships - - - L51 XVI. Strains experienced by Ships - - 154 Index ... - ... loi With 18 Plates. S 11276—1250—2/91 Wt 22921 D & S. A 2 NAYAL ARGHITEGTURE. CHAPTER I. Buoyancy of Ships. By luoyancy is meant the upward support given to a ship by .the water in which she is immersed. The submerged part of a vessel at rest in still water is subjected to fluid pressure, which acts, at each point, in a direction perpen- dicular to the surface of the ship at that point. The pressure on .any one part may — by the principle of the resolution of forces — be supposed replaced by three others having the same effect upon the ship ; one acting vertically, the second horizontally athwart- ships, and the third horizontally fore-and-aft. To readers unac- quainted with the above principle this may be made somewhat •clearer by an illustration. Imagine a body, w |^ denoted by P in Fig. 1, acted upon by three forces X, T, and Z, along three strings mutually at right angles (like the edges of a box at one corner) ; these will tend to make the body move in a certain direction, which movement may be prevented by applying a force (W, say) by means of a string in a direction opposite to that in which the body tends to move. Thus the effect of the three forces X, Y, Z, is equal and opposite to that of the force W, and the three first-named forces may therefore be replaced by a single force F equal to W and acting in the opposite direction. Conversely, if the force F operated upon the body P, it might be supposed replaced by the three others X, Y, and Z, which have the same effect. "We may therefore conceive the pressure upon each part of the ahip replaced by three others, thus obtaining three sets of parallel pressures, one set acting vertically, another horizontally athwart- ships, and the third fore-and-aft ; noticing that some of the pressures in any one set may be acting in the opposite direction to others in the same set. The ship being supposed at rest, there is no motion in an athwartship or fore-and-aft direction, and hence Fjc.Z. — \ . J :i p\ the sets of pressures in those directions must respectively balance amongst themselves ; that is, the sum of the pressures, P in Fig. 2, on one side of the ship must be equal to the sum of the pressures- on the other side ; and the total pressure Q urging the ship ahead must be equal to that tending to make the vessel go astei'n. The value of P is about 3,500 tons in a large ship, and of Q between 400 and 500 tons. The vertical fluid pressures are the most important, as these give the huoyancij which supports the ship. Since any number of parallel forces may be supposed replaced by a single force — or resultant as it is termed — acting parallel to them and equal to their sum, it is only necessary to determine the magnitude of this resultant and the position of its line of action, in order to find the effect of all the vertical pressures upon the ship. Before- proceeding to do this, one or two definitions will be given. The centre of gravity of a body may be defined as the point through which the weight of the body, when at rest, may be supposed to act, in a direction vertically downwards. When a ship is immersed in water, it occupies a cei-tain space which would otherwise be filled with water ; the quantity of fluid so dislodged is called the displacement, and may be measured by its volume in cubic feet or its weight in tons ; in the former case it is known as the volume, and in the latter the weight of displace- ment. "When " displacement " alone is spoken of, the weight of displacement is usually meant. The magnitude, and position of the line of action, of the resultant vertical pressure may now be determined by the aid of the following : — Proposition : When a vessel is immersed in water in any position, it experiences an upward pressure which is equal to the Aveight of the displacement in that position ; and this pressure acts through the centre of gravity of the displacement. Proof .- In Fig. 3, suppose the vessel A C D held forcibly in the position shown. Imagine the water surrounding this vessel solidified and that then the latter is removed, thus leaving a cavity Fic.3. having a volume equal to that of the displace- ment. If now this cavity be filled with water of the same kind :rz~ as that in which the -— — vessel was immersed, and the solidified water again becomes — liquid, there will =^ obviously be no disturbance of level, and the same pressures which were before acting upon the immersed portion of the ship are now acting upon and supporting a mass of water having a weight equal to the displacement of the vessel. Now the weight of displacement acts vertically down- wards through its centre of gravity, marked B, and therefore, to support this weight, the upward fluid pressure must be equal to the weight of displacement, and its line of action must pass vertically through the point B, since in order that two equal forces may balance each other they must act in the same straight line. Hence also the upward fluid pressure (or buoyancy) on the sMp must be equal to the weight of water displaced, and must act through the centre of gravity of the displacement ; this point is called the centre of luoyancy. The situation of this point when the vessel is floating in any assigned position can be calculated and the line of action of the buoyancy thus determined. When a vessel is upright athwartships, its position in the water is defined by the depth or draft of water at the fore and after perpendiculars (between which the length of the ship is measured) to the underside of keel, or that produced ; e.g., ah in Fig. 4 is F/C.4. E=^^ d r 6 ^he draft of water forward, and cd is the draft aft. The mean draft of water is that midway between the perpendiculars, as ef, and is evidently half the sum of the drafts forward and aft. The difference between the extreme drafts is known as the trim of the ship, and when the excess draft is aft, as is usually the case, the vessel is said to trim by the stern. The line in which the surface of the water cuts the surface of the ship when floating in any position is called the water line for that position, the area enclosed by that line being the water i^lane area. The line corresponding to the fully laden condition of the ship is the lo€ud water line. In the case of a vessel floating freely and at rest in still "water, the only forces acting vertically are the weight of the ship down- wards through its centre of gravity, and the buoyancy upwards through the centre of buoyancy ; and since these balance, the following conditions must hold : — I. The weight of the ship must be equal to the weight of water displaced. II. The centre of gravity of the ship and the centre of buoyancy must be in the same vertical line. A ship being desired which shall have certain weights of armour, guns, &e., an estimate can be made of the weight of the structure (or hull) necessary to carry the above safely at sea. The total displacement — being the sum of the weights carried and structure to carry them — can thus be ascertained, as well as the position of the centre of gravity of the ship. Then the second condition of rest mentioned above, teaches that the under-water part of the hull must be so shaped as to bring the centre of buoy- ancy in a line with the centre of gravity, whilst the first condition enables the size of this part of the hull to be determined, since if the weight of displacement is known, its volume can be calculated when the weight of a cubic foot of the water in which the vessel is to be immersed is known. Sea water, in which war ships principally float, weighs 64 lbs. per cubic foot, so that 35 cubic feet weigh one ton. If therefore the weight of displacement is multiplied by 35, the required volume of displacement in sea water is obtained. Conversely, dividing a known volume of dis- placement of a ship in sea water by 35 will give its weight. To take a simple example : a box-shaped vessel 210 feet long and 40 feet broad floats in sea water at a draft of 15 feet forward and aft ; what is its weight ? '^^placTSis"' '":} =210x40x15=126,000 cubic feet. ""dis^^latmlnP* °- } =126,000^35=3,600 tons. To get the volume of displacement of a ship-shaped vessel when floating at a given water line, the drawings showing her form and size are necessary, the volume being calculated from them by the aid of well-known rules. Thence the weight of displace- ment can be obtained. Fresh, and river water difEer in weigM from sea water. The former weighs 62^ lbs. per cubic foot, whilst the latter varies in weight according to its position relatively to the mouth of the river ; that at the London Docks weighs 63 lbs. per cubic foot. In consequence of this, a ship will sink in the water on passing from the sea into a river, since a greater volume of displacement .,is necessary in the latter case to make up the same weight of ship. This sinkage is about 3^ inches for a ship of 10,000 tons displace- ment, and is, conversely, the amount she would rise on passing from a river to the sea. As the stores, &c., of a ship are consumed, the displacement becomes less, and the ship therefore rises in the water ; e.g. FiG.S. SCALE OF D/SPLAC£MENr IN TONS. X eooo sooo 1OO0 3000 zooo AMIDSHIPS. 10 the Inflexible rises 27 inches due to the consumption of lier coals aione. It is thus often requisite to know the displacement up to water lines below that for the load condition. To ascertain this without the labour of direct calculation, a " curve of displace- ment " is used, which is constructed in the following manner : — A vertical line AX, Fig. 5, is taken, on which to set ofE — usually on a scale of half an inch to a foot — mean drafts, measuring from A, which represents the underside of keel amidships, or zero mean draft. Let AB denote, on the proper scale, the mean draft corresponding to the load displacement. Draw BBj perpendicular to AX and let its length represent the load displacement on a certain scale, a quarter of an inch often representing 200 tons dis- placement for a large ship. The same construction is made for several other water lines parallel to the load water line, the ship being supposed to lighten without change of trim. Let AC, AD, &c., denote the various mean drafts corresponding to these water lines, and CC^, DD^, &c., the displacements at those drafts, obtained by direct calculation from the drawings. A curved line drawn through the points, B^, Cj, Dj, &c., is the "curve of displacement,"^ its use being to ascertain, by a simple measurement, the displace- ment for any draft between those for which it has been actually calculated. Suppose, for example, that the particular ship of which the curve is given in Fig. 5 floated at a mean draft of 20| feet on a certain occasion. To ascertain the corresponding displacement at that time it is only necessary to set off A.in to represent 20-^- feet on the proper scale, and to draw mm-^ perpendicular to AX to meet the curve in m^. Then min^^, which represent 6,600 tons on the proper scale, gives the displacement required. It should be noticed that this supposes the ship to have the same trim at the mean draft of 20| feet as in the load condition, the cur-^e having been constructed on that assumption ; but the displacements obtained from the curve are also very approximately correct for all but extreme departures from that trim. Ships have a form of displacement curve similar to that shown in Fig. 5, but a particularly simple case is that of a box floating evenly in the water, having the same draft forward and aft. It is evident that here the " curve " becomes a straight line, for at one-half the load draft the displacement is one-half that in the load condition ; at one-third the load draft the displacement is- one-third the load displacement, and so on, the displacement being proportional to the draft. The converse case, viz., to determine the draft corresponding to a given displacement can also be easily dealt with ; for it is only 11 necessary to draw a straight line parallel to the base line of the- curve, at a distance representing the given displacement from it^ and the draft corresponding to the point where this line meets the curve is the draft required. In this way also could be found out the sinkage due to putting weights of moderate amount into the ship when floating at any assigned water line ; for the draft corresponding to the new displacement can be ascertained as ■above, and the difference between this and the original water line will be the sinkage sought. This problem can, however, be more easily solved by finding the weight necessary to sink the ship one- inch from the assigned water line — known as the " tons per inch immersion" at that line— and dividing the given weighf by the quantity so found, thus obtaining the consequent sinkage in inches. It is here assumed that the weight necessary to sink the ship one inch from the assigned line is equal to that which will immerse her each additional inch until the total sinkage is reached, an assumption which is very approximately true for- moderate changes of draft within the limits at which the ship is likely to be floating on service. To FIND THE " TONS PER INCH " AT ANY WATER LiNE : The weight which will make the ship sink one inch without change of trim must obviously be equal to the weight of the added displace- ment, because of the necessary equality between the weight of the ship and the weight of displacement ; thus, if A represents the area of the water plane in square feet, the added displace- ment due to sinking one inch must be A x ^V cubic feet ; i.e., A X iV H- 35 = j^ tons. Hence we get this rule : — Tons per inch ] immersion at } _ Area of water plane at that line in square feet any water line ) 420. This also gives the weight which must be removed to lighten the ship one inch from the same water line. The areas of several water lines parallel to the load line are determined from the drawings, and the " tons per inch " at those lines obtained, the results being set off on a " curve of tons per inch immersion " as shown in Fig. 6. The horizontal measurement at any draft denotes- the " tons per inch " at that draft, a quarter of an inch very often representing one ton in the case of a large ship. The use of this- curve is that it gives, by a simple measurement, the "tons per- 12 F/G. 6. SCALE OF TONS PER /NCH. Inch " at drafts between those for which it has been actually -calculated. Suppose, for example, it is desired to know how much the ship having the curve of tons per inch shown in Fig. G will sink if 400 tons of coal are put on board when she is floating .at a mean draft of 20 feet : set up Km to represent 20 feet on the proper scale, and then mm-^ will give the tons per inch — 33^ — at that draft. Hence — Sinkage required = qrn; = 12 inches very nearly. An examination of Fig. 6 shows that in the vicinity of the load water line the curve is nearly parallel to the base line, thus justifying the assumption previously mentioned that for moderate changes of draft hear the load line the " tons per inch " is practically constant. 13 Reverting to the box, the displacement of which has beam previously found, it is seen that the area of the water plane- ia the same at all drafts — supposing no change of trim — viz.: 210 X 40 square feet. Thus the "tons per inch" is 210 X 40 -f- 420 = 20, and the " curve " of tons per inch is a straight line parallel to the base. When the drawings of a ship are not available from which to- calculate the area of the load water plane, a good approximation to the " tons per inch " for that plane may be made by rules- given below ; these have been taken by permission of "W. H. White, Esq., F.R.S., Assistant Controller of the Navy and Director of Naval Construction, from the " Manual of Naval Architecture." Let L be the length of the ship in feet at the load water line, and B the extreme breadth at that line ; then, if the water line were rectangular, its area in square feet would be given by the product L x B, and the- " tons per inch " by L x B divided by 420. The actual area of the water plane is, however, less than that of the rectangle, being; Fic.7. < ... L > ^ '- """^^^^^^^ "fined" down from the latter at the ends, as shown in Fig. 7, so- that the tons per inch for the ship will be obtained by dividing the area of the rectangle by some number greater than 420, its magnitude depending upon the extent to which the plane is- " fined " at the ends. If the product L x B be represented by A, the following rules give, approximately, the tons per inch at the load water line of ships having difiEerent degrees of fineness at the ends : — Tons per Inch. 1. For sMps with fine ends = i y^ ^_ 2. For ships of ordinary form, (including probably the ) __ i . great majority of vessels) j 6<>u ^ • 3. For ships of great beam in proportion to the length, I i „ , and ships with bluff ends ( 'S'^'^ ^ ■^• In many ships, L is very nearly indeed equal to the length of the ship between the fore and after perpendiculars, in others being a little longer ; whilst B is usually equal to the extreme breadth of the ship : in these approximate calculations, therefore, no great error will be involved if the length between perpendiculars and the breadth extreme of the ship are used as the values of L and B respectively. 14 To illustrate the application of the above rules, suppose the tons per inch at the load line of the Orlando to be required ; here L=300 feet, and B=56 feet, and as this vessel has fine ends, the first rule will apply ; hence Approximate tons per inch=- — -^^ — =28 tons. The actual tons per inch, as obtained from the drawing, is 27^ tons, so that the approximation is good. As a ship of ordinary form, the Invincible may be taken, having L=280 feet and B=54 feet, so that Approximate tons per inch='^- v^,.' =27 tons. This being also the value as obtained from the drawings. If any damage occurs to the under water part of a ship, the displacement becomes less than before, due to the entry of water ; "the A-essel therefore begins to sink, and will almost inevitably founder unless structural or other means are provided to limit the accumulation of water. This may be done by employing pumps to eject the water, thus keeping the leak under ; or some form of " leak stopper " may be used to prevent the ingress of water, such as the " shot hole stopper mat " for holes near the water line, or " collision mats," sails, etc., in the event of a more ■dangerous leak below water ; or, finally, the interior of the vessel may be divided into numerous watertight compartments by vertical and horizontal partitions, thus restricting the water to the particular compartment in the neighbourhood of the damage. As to the first method, it will be shown (Chapter XI.) that the pumping power supplied to a ship is totally inadequate to the task of keeping down a large leak, such as might occur from the •explosion of a torpedo, or the assault of a ram ; whilst as to the second, it is extremely difficult to get a leak stopper into position over a jagged hole, through which water is continuously and rapidly flowing. Efficient watertight subdivision must therefore be adopted in all ships, as it must be mainly relied upon to secure the safety of a vessel when greatly damaged below water, the pumps and leak stoppers being regarded as auxiliaries. Details of this subdivision will be found in Chapter X. If certain compartments of a ship so divided become damaged and filled with water, the buoyancy due to the displacement of those compartments is lost, and the ship will sink until by her extra immersion an additional amount of watertight volume (previously above water) has been put into the water to restore the necessary equality between the weight of the ship and the weight of displace- 15 ment. Th.e total volume, and corresponding buoyancy, of the ■watertight space lying above the water line is called th.e reser ve_of buoyancy, as distinguished from the volume of the ship helow that line, which measures the buoyancy utilized. This reserve of buoy- ancy usually includes the volume of the ship lying between the water line and upper deck, together with any superstructures, poop, forecastle, etc., which can be closed in and made thoroughly watertight.. The preceding has a most important bearing upon the safety of the ship, not only as regards foundering — by being in reserve should a loss of buoyancy occur through damage below water — ^but also, as will be shown in Chapter II., in respect to her immunity from being capsized. It is usually expressed as a per- centage of the displacement, and varies from 10 or 20 per cent, in the low sided American monitors to more than 100 per cent, in high sided vessels with very fine under water forms. To illustrate the foregoing, and the general method of ascer- taining how much a ship will sink when certain compartments are damaged below water, the following example may be taken : — A box-shaped vessel, 280 feet long, 50 feet broad and 27 feet in total depth, floats at a draft of 18 feet forward and aft, and has a central compartment 35 feet long, bounded by two watertight athwartship partitions the whole depth of the box. Find the reserve of buoyancy ; and the distance the vessel would sink if the central compartment — supposed empty — were laid open to the sea. F/Q.8. , A) Bt . . •w, V//////////A V///////////////////^////////////777. '//////////////77777/^ A B In Fig. 8, let AA^ and BBj represent the two partitions, or "bulkheads," and WL the water line at which the box floats before damage. Then, since the volume above water is obviously one-half that below it, the reserve of buoyancy before damage is 50 per cent, of the displacement. Siappose now that the compartment becomes injured, and that the vessel is forcibly held at the water line WL until the water rises to its own level in the compartment. The part ABCD, which before displaced its own volume of water, no longer does so, and the volume of displacement thereby lost is 35 x 18 x 50:= 31,500 cubic feet. Thus the box, on being released, must sink 16 bodily in the water without change of trim (the compartment being central) until this lost buoyancy is restored. Let Wj L^ be- the new water line when this is the case, the water consequently rising in the compartment to the height EF. It is thus seen that the part of the reserve of buoyancy lying between AA^ and BB^ is also lost, since it does not displace water as the vessel sinks ; and the ends alone must be looked to in order to supply the lost displacement ; that is, the volume "WjEDW, together with the volume PLiLC must be equal to 31,500 cubic feet. Now the sum of these volumes is obtained by multiplying the intact area of the water plane {i.e., the total area of the plane excluding the part CD) by the sinkage of the ship (WWi) in feet : this intact area is (280 — 35) X 50 = 12,250 square feet, so that Sinkage of the vessel in f eet= Loss of buoyancy in cubic feet Area in sqn are feet of intact water plane- _31,500_2,„ ~I2;250-'^^*^^*- If the compartment had been partially filled with machinery, stores, &c., these would have contributed their own volume of displacement after the ship had been damaged, and the loss of buoyancy would have been less than before by an amount equal to the volume of these stores : in other words, the loss of buoyancy is- always that due to the unoccupied space in the compartment, up to the height of the water line WL. Suppose, for example, that one-fourth the space ABCD had been occupied by stores, the- loss of buoyancy, and consequently the sinkage, would only have been three-fourths the amount calculated above. Thus,, where convenient, stores such as coal, &c., should be placed in those parts of the ship most liable to damage, as in the region of the water line at the unarmoured ends of central citadel ships- {see Plate X). The method used in the example above to ascertain the sinkage- is also employed for ship-shaped vessels, the procedure being to first ascertain the unoccupied space in the damaged compartment up to the original water line, and then the area of that part of the water plane which is unaffected by the damage ; the division of the former by the latter gives the sinkage required. If the compartment is not central, so that change of trim occurs, the above calculation for the sinkage must be combined with an estimate of the change of trim, in order to get the final position of the ship after sinking. It has been noticed above that the part of the reserve of buoyancy lying between the athwartship bulkheads was'lost when. 17 the compartment became damaged ; this may be avoided by fitting' a horizontal watertight partition or " platform " at the height CD, thus preventing the -water from rising in the compartment above that height as the vessel sinks ; and as the whole of the water plane area is now "intact" instead of only seven-eighths as before, the consequent sinkage will only be seven-eighths that previously calculated. Thus the protective watertight decks, worked in many ships approximately at the height of the water line, are very valuable for protecting the reserve of buoyancy in the event of damage below water, as well as for preserving the buoyancy from the effects of gun fire. If the platform had been worked, say, 9 feet below water instead of at C D, and the leak occurred below it, the loss of buoyancy would have been only one-half what it was before, since water could not rise above the platform ; and as in this case also the total water plane area would be intact, the sinkage would only have been one-half that in the preceding case. If the damage had occurred above the platform, the loss of buoyancy would have been the same as when the leak was below it ; but as only seven-eighths of the water plane would then be intact, the sinkage would have been greater than in that case, although only one-half as much as when no platform at all is worked in the compartment. It is thus seen that the uses of horizontal watertight platforms below water in case of accident are to lessen the loss of buoyancy, or both the loss of buoyancy and water plane area, according as they are below or above the leak. Finally, it is seen in the above example that if the athwartship bulkheads were carried to a less height in the ship than E F, water would flow over their tops, as the vessel sank on being damaged, into adjacent, compartments — imless the bulkheads were terminated at a watertight platform — and would therefore be of little or no use. Hence the deduction that all the main vertical ' watertight partitions in a ship should either be carried sufficiently high to ensure that their tops would still be above water after any sinkage, change of trim, or change of heel of the ship which is likely to occur from damage in action ; or their tops should stop at a watertight deck or platform. Examples. I. A box-shaped vessel 70 feet long and 20 feet broad has a total depth of 13 feet, and a draft of 10 feet forward and aft in sea water. Express the reserve of buoyancy as a percentage of the displacement. Also, find the displacement in cubic feet and in tons, and the tons per inch immersion. S 11276 B 18 Answers — 30 per cent. ; 14,000 cubic feet ; 400 tons ; 3^ tons per inch. II. The water lines of a vessel are 4 feet apart, and the dis- placements up to these lines are respectively 7,800, 6,240, 4,700, 3,360, 2,040, 880, tons, commencing from the load water line. Construct to scale the curve of displacement, and find the displacement corresponding to 18 feet mean draft. Answer — 5,480 tons. III. The tons per inch immersion at the water lines mentioned in the preceding question are 32"6, 32'3, 31'4, 30*1, 27'7, 23"0, tons respectively. Construct to scale the curve of tons per inch immersion, and find the tons per inch at 18 feet mean draft. Answer — 31*9 tons. IV. The vessel in question I. has a central compartment 10 feet long, bounded by watertight bulkheads which extend from top to bottom. If this compartment is laid open to the sea, how far will the vessel sink ? Ansiver — 20 inches. 19 CHAPTER II. STABILITY OF SHIPS. A SHIP'S stability is that quality by virtue of whicli she tends to right herself ^hen inclined from her position of rest. In the preceding chapter it was shown that when the displace- ment of a ship is equal to its weight, and the centres of gravity and buoyancy are in the same vertical line, the vessel will be at rest '(or in equilibrium) in perfectly still water, if subject to no ex- 1;ernal disturbing causes. In practice, however, this freedom from •disturbance cannot be secured, so it becomes important to investi- gate the behaviour of the vessel when displaced from her position of rest by external forces such as the wind ; in other words the ■character of the equilibrium is required. If a ship when slightly inclined in any particular direction from her position of rest re- turns towards that position when the inclining forces are removed, it is said to be in stable equilibrium for the given direction of in- clination ; if on being released it moves farther from its position •of rest it is in unstable equilibrium, whilst if it remains in the slightly displaced position, without any tendency to return to- ■wards, or move farther from, its original position of rest, it is in .neutral or indifferent equilibrium for the given direction of inclination. As an example of a body in stable equilibrium, reference may be made to a pendulum with the bob vertically beneath the point of suspension ; the same pendulum with the bob balanced vertically above the point of suspension is in unstable ■equilibrium, and a uniformly heavy sphere on a horizontal surface furnishes an illustration of a bo'dy in neutral equilibrium. In proceeding to discover the character of the equilibrium of a iship, it may be inclined in either a transverse or longitudinal direction, or in directions between those extremes ; but since the former is by far the most important, it is to inclinations in a transverse direction that attention will be chiefly directed, the «hip being supposed to maintain the same volume of displace- ment throughout the inclination. S 11276 B 2 Let Fig. 9 represent a ship floating freely and at rest in still water at the water line WjLi,G marking the height of the centre- of gravity and B the position of the centre of buoyancy, G and B- being in the same Yertical line. Suppose now the vessel is slightly- displaced transversely (through 1° or 2°, say) and forcibly held in. the position shown in Pig. 10, W2 Lj being the new water line. If no weights on board have shifted during the inclination,, the centre of gravity is in the same position in the ship as- before, but since the shajpe of the volume of displacement has altered, the centre of buoyancy will now be at some point B^ out of the middle line of the ship ; and the line of action of th& buoyancy will be that drawn vertically upwards through Bj^.. This line of action may, or may not, pass through G. If it does, the- two conditions of equilibrium, are satisfied after the ship is in- clined, and she will therefore remain at rest in that position ; that, is, the original position of rest is one of neutral equilibrium. Generally, however, the line of action of the buoyancy will not pass through G, but will cut the originally vertical line B G in some point M either above or below G. Suppose it to cut above G, as in Fig. 10 ; then the weight of the ship (W tons, say) acting downwards through G, and the buoyancy upwards throiigh Bj, form two equal and opposite forces acting on the ship ; that is- they constitute a " mechanical couple," the effect of which would, be to turn the vessel towards her. upright position if the external forces were removed which are forcibly inclining the ship. Thus when M is above G, the ship is in ^taUe equilibrium. On the- other hand, if M fell lelow G, the weight and buoyancy would form an upsetting couple, and the ship would be in unstable equi- librium when upright. It is therefore seen that the character of the- 21 •equilibrium depends entirely upon the position of G relatively to Ihe point M. This latter point is called the metacentre, probably because in a given vessel it is the " meta " (or limit) beyond which .the centre of gr3,vity may not rise for stable equilibrium. All .ships should be stable when ready for sea, and hence the designers must adopt nieans to ensure that G falls below M in that condition. Now the vertical distribution of the weights of a ship, such as ^lins, armour, &c., determines the height of the centre of gravity, -and since this distribution is practically fixed by given conditions in the design, such as the height at which guns and armour should be carried at sea, little control can be exercised over the vertical position of the centre of gravity. By altering the shape of the ship at the load line, however, and varying the under-water form, the metacentre can be made to occupy very different positions in the vessel whilst maintaining the same volume of displacement ; it is high up in broad, shallow ships, and low down in narrow, ■deep ships. Hence, the naval architect, after calculating the position of the centre of gravity of a new ship, adopts such a form at the load water line, and under water, as will make the meta- centre fall above G. The position of the metacentre in a given •ship is obtained by calculating, from the drawings, its distance from the centre of buoyancy of the vessel when upright, (which distance for a given volume of displacement is entirely dependent upon the size and shape of the water plane area) the actual position of that centre of buoyancy being also obtained from the same source. The found value of BM being set off from the known position of B will give the point occupied by M in the ship. This marks the point, in the originally vertical line, through which the line of action of the buoyancy passes when the ship is slightly inclined from the vertical ; but, in ships of usual form, this holds good, approximately, for angles up to 10^ or 15°. Hence, when once the position of M has been determined, the line of action of the buoyancy for any given angle less than the above can be obtained by drawing through M a straight line making the given angle with the vertical. The distance between G and M is called the transverse meta- ■centric height or, shortly, the metacentric height : its value is very •different in different classes of vessels, and has to be adjusted to meet conflicting claims. For example, in order that the ship may be stiff— i.e., difficult to incline by external forces such as wind pressure on sails^ — the metacentric height should be great ; whereas to ensure the ship forming a steady gun platform at sea, GM must be made small. The second statement is shown to be true in Chapter III. ; the truth of the first can be demonstrated as follows : — From G in Fig. 10, draw GZ perpendicular to BiM ; GZ is the arm or lever of the mechanical couple formed by the weight and buoyancy ;; and as the turning effect of any couple is measured by its moment (obtained by multiplying either force into the perpendicular distance between them), the effort which the ship makes to right herself is W x GZ foot tons, W being expressed in tons and the length of the arm GZ in feet. This effort of the ship to right herself when inclined at any angle, is called her statical staUlitif at that angle, and must be equal and opposite to the effort of the- external forces keeping her inclined ; and since to counteract the turning effect of a couple another couple of equal moment must be employed, it follows that the external forces necessary to hold a ship at a given angle must form a couple, the moment of which is equal to the statical stability at that angle. Suppose, for example, the ship is inclined by a wind pressure of P tons acting on the- sails ; this will cause the ship to have leeway until a corresponding pressure P acts upon her under water surface, after which a uniform motion to leeward will be established, and the vessel will remain at a steady angle of heel. Let the forces P act as shown in Fig. 10,. the perpendicular distance between them being h feet. Then Pxh=WxGZ Now, for the moderate angles of heel usually experienced by ships'- under wind pressure in smooth water, the line of action of the buoyancy passes through the metacentre, so that GZ = GM sin. 9, where is the inclination in degrees. Thus we may write Pxh=:WxGM sin. 9 This equation enables the angle of heel to be determined, to which a given wind pressure will incline a particular ship : e.g., let W=8000 tons, GM = 3^ feet, P = 14 tons and A = 100 feet ; then 14x100 1 Sin. e-- 8000 x3i 20 Whence, from a table of sines, = 3 degrees nearly. The above- equation also shows that in a ship of given weight, the greater the- metacentric height (GM) is, the greater will be the moment of the- external forces (Pxh) necessary to hold the ship at a given angle 9 ; or, in other words, the greater the metacentric height is- made, the smaller will be the angle to which an external couple of given moment will incline the vessel. Hence, as stated above,, to ensure that the ship shall not be easily inclined by external forces, the metacentric height should be made as great as possible- Seeing, then, that for stifEness the metacentric height should be great, and for steadiness in a seaway small, the GM actually given 23 is mainly determined by experience with successful vessels ; such a form of ship being adopted as will make the metacentric height small enough to secure steadiness, without undue sacrifice of stiff- nees ; observing that the special forms adopted in certain vessels from other considerations — such as in central citadel armour-clads to provide against the reduction of metacentric height which would occur, in the probable event of their ends at the water line being damaged in action ; or in armoured coast defence vessels where moderate draft is necessary — often lead to metacentric heights greater than are required from considerations of stifEness alone. The following table shows the values given in various iniportant classes of vessels : — Name and description of ship. Metaoentrio Height. 1. JS'orthumherland : early broadside armour-clad 2. Devastation : Mastless turret ship 3. Inflexible : do. (central citadel) 4. Monarch : masted turret ship 5. Glutton : coast defence vessel 6. Medina : gunboat for river service ... 7. SealarJi : sailing brig 8. Iveiinstant : unarmoured cruiser 9. Modem protected cruisers, having steadying sails only 10. Seagoing gun boats and gun vessels 11. Torpedo boats 1st class . 12. Do. 2nd class Feet 3} 8i 2* 7 12i 2i About 2 2ito3 li to2 About 1 A comparison of the Sealark and Medina well exhibits the influ- ence of form upon the metacentric height. They have practically the same displacement (380 tons for the former against 363 tons for the latter), but whilst the Sealark has a breadth of 29J feet and a mean draft of 13 feet, the Medina is 34 feet broad and has a mean draft of only 6 feet ; thus M is much higher in the latter ship (see page 21), which conduces to a greater metacentric height. The above values of GM are those which the vessels have when they are fully laden. But as a ship con- sumes her stores, coals, &c., not only does she float at lighter drafts, but the positions of the meta- centre and centre of gravity become altered. It is there- fore important to see F/0. II. 24 that the vessel is sufficiently stiff in the various lightened condi- tions. Suppose WL is the load-water line of the vessel shown in Fig. 11, B and M being the corresponding centre of buoyancy and metacentre respectively. Also, suppose that after certain stores are consumed the water line is at W^Lj. The centre of buoyancy (Bj) corresponding to W^Lj can be ascertained from the drawings, as also the distance (B^Mi) between this point and the new meta- centre Ml. Thus the position of the metacentre in the lightened condition may be set off, and will not, generally speaking, coincide with M. The different positions of the metacentre for the several lightened conditions are usually shown on a metacentric diagram, which is constructed as follows : — F/G. 12 i^ M, rn M A horizontal line WL, Fig. 12, is taken to represent the load water line, and through a convenient point L in it a straight line LLj is drawn at an angle of 4.5 degrees to WL. A series of equidistant lines W^L^jWjLj, &c., are then drawn parallel to WL at the proper distance apart (usually on a scale of half-an-inch to a foot) to represent successive water lines at which the ship floats as she is lightened, it being su^jposed that the ship rises bodily without change of trim. Through the points L, Lj, Lj, &c., where the water lines meet the inclined line L Lg, vertical lines are drawn, and on them are set off the distances LB, L^Bj, LjBj, &c., to represent (on the same scale as above) the depths of the respective centres of baoyancy below their corresponding 25 water lines, these distances having been previously calculated from the drawings. The distances B M, B.i M^ Bj Mj, &c. between each centre of buoyancy and corresponding metacentre are like- wise calculated, and set up from the points B, B^, Bj, &c. A curve passed through the centres of buoyancy is called the curve of ■centres of buoyancy, that through the points M, Mj, Mj, &c. being knovTn as the curve of metacentres. It will be noticed that the line L L3 was drawn at an angle of 45 degrees in order to have the vertical lines in the diagram the same distance apart as the horizontal water lines. On the left hand side of the diagram a table is usually arranged, giving particulars of the displacement &c. at each water line. The use of this diagram is that by its means the positions of the centre of buoyancy and metacentre for any draft between those for which the positions of the centres of buoyancy and metacentres have been actually calculated, can be easily ascertained. Let tu I be some intermediate water line for which these particulars are required. From I, where w I meets L Lg, draw a vertical line to meet the curves of metacentres and centres of buoyancy in m and b respectively. Then the metacentre, when the ship is floating at the water line represented by w I, is at a distance I m (on the proper scale) above that line, and the centre of buoyancy is at a distance represented hj lb below it, presuming of course that the ship has the Bara.e trim as when floating at the load water line. Having a metacentric diagram given for a particular ship, the metacentric height for any lightened condition can be ascertained when the position of the centre of gravity of the vessel in that lightened condition is known. This latter point can be found in the following manner, if (as is usual) the centre of gravity of the vessel in the load condition is known, and the weight and position of the stores to be removed to bring the ship from the load to the particular lightened condition are given. Let G, in Fig. 11, denote the position of the centre of . gravity when the ship is fully laden ; and let w represent the weight in tons which has to be removed in order that the vessel may float at the water line Wj L^. Then if the centre of gravity of this weight before removal happens to coincide with G, the ■ centre of gravity of the ship after the weight is removed will still be at G, and the metacentric height in the lightened condition will be GM^. If, however, the centre of gravity of the weight to be removed is h feet below G, the centre of gravity of the ship will rise from G to G^ where GG,=.^^feet, 26 W being the weight of the ship in tons when fully laden. Then the metacentric height in the lightened condition is G^Mi. On the other hand, if the centre of gravity of the weight iv is h feet above G, the centre of gravity of the ship when lightened will fall below G a distance given by the above formula. If stores of total weight lu tons are add-ed to a ship of weight W at a distance of h feet above or below its original centre of gravity, the centre of gravity of the ship after the stores are put on board will be above or below its original position a distance of Trr feet. The particular condition to which attention is mainly directed is that when all the consumable stores, such as coals, water in boilers, provisions and officers' stores are out of the ship, this being known as the " light " condition. The metacentric height for this condition is shown on the metacentric diagram of a ship,, in addition to that for the load condition. To take a numerical example, suppose the weight of consumable stores on board the ship, the metacentric diagram of which is shown in Fig. 12, is 1,4.50 tons ; also that the centre of gravity of this weight is 2"1 feet below the centre of gravity of the ship in the fully laden condi- tion, and that the metacentric height for this condition is 3 feet, so that G M in Fig. 12 represents 3 feet. It is required to find the- metacentric height in the " light " condition. When the weights are consumed, since there are 1,100 tons displacement between WL and WjLj^, the water line for the light condition will be 350 .-j^3=ll inches nearly below W, L^, the tons per inch at "W^jL^ being- 32-3. Let WL' represent the water line for this light condition. Then M' will be the position of the corresponding metacentre. Again, the rise in the centre of gravity due to the consumption of stores is 1450 X 2-1 3045 - „ ^ 7600-1450 = 6l50 = "'^^""'^"^"^y- Thus, in setting off the vertical position of G', the centre of gravity in the light condition, it will be 6 inches nearer to WL than G is. Then G' M' represents, on the proper scale, 2-6 feet,, which is therefore the required metacentric height in the light condition. The metacentric height of a war ship in the light condition is- usually. from 6 inches to 1 foot less than its value in the load condition, although in some ships the decrease is greater than this, and in others less. In merchant ships the variation between the 27 load and light conditions is greater than in war ships, and it not unfrequently happens that a vessel which has a good metacentric height in the load condition has a negative metacentric height in: the extreme light condition. That is to say, in that extreme- condition the centre of gravity lies above the metacentre, so that the ship is in unstable equilibrium. Such a ship will not remain upright, but will loll over several degrees until it reaches- a position of stable equilibrium. Of course this will only occur in harbour, after discharging cargo, and a ship having this character- istic may be rendered stable in the upright position by the introduction of rubble or water ballast low down in the ship,, which considerably lowers the centre of gravity without appre- ciably affecting the position of the metacentre. f/a. /3. Having the metacentric heights in the extreme light and load conditions, the question arises as to what these heights will be for intermediate condi- tions of lading. These will depend upon the order in which the stores are consumed. If stores lying high up in the ship are first used, the centre of gravity will first begin to fall, and in going from G in the load to G' in the light condition, the varying positions of the centre of gravity will be some- what as represented by the curve GaG' in Fig. 13. In this case the m.etacentric height is least in the light condition. If, however, the lower stores are first consumed and afterwards any which are very high up, the movement of the centre of gravity will be appioximately as shown by the- curve G 6 G', in which case there will be some intermediate con dition of lading for which the metacentric height will be less than in either the load or extreme light condition. In most war ships,, nearly all the consumable stores are below the centre of gravity for the light condition, so that in whatever order they are consumer! the centre of gravity of the ship for intermediate conditions of lading- can rise but very little above the centre of gravity for the light condition ; hence the successive positions of the centre of gravity- will be approximately as shown by the curve G c G'. Thus- it follows that in the majority of war ships, since the curve MM' is fairly representative of the curve of metacentres of such ships, it may reasonably be concluded that if sufficient stiffness is secured 28 in the light condition, there will be enough in all other conditions. In those ships, however, which have upper coal bunkers, the movement of the centre of gravity will be of the character shown by G 6 G', if the coal and stores are worked out from below before taking coal from those bunkers. This is the more important in •■ships of moderate dimensions having relatively large stowage in the upper bunkers and a not very great metacentric height (2 ft. say) in the load condition. In such ships it is not advisable to •completely empty the lower bunkers before trimming coal from the upper ones. Not only is the position of the centre of gravity of a ship .altered by the consumption or addition of stores, but also by every movement of weights on board. The amount of the change can be estimated by means of the following general mechanical principle : If of a body, of total weight W, a part of weight lu is moved through a distance of h feet in any direction, the centre of gravity of the whole will move parallel to that direction .through a distance of Let A B C D E in Fig. 14 represent any body of weight W. Suppose the centre of gravity of the part ABE (of weight lu) is at a, and that of the part B E D C at &, the centre of gravity of the whole body being at G. If, now, the part A B E is moved, so that its centre of gravity is taken from a to c, a distance of h feet, the centre of gravity of the whole, after this movement, will be at G^ on the line joining h and c, where G Gj is parallel to a c and equal in length to — =r-feet. The proof of this depends upon the fact that the common centre of gravity of two bodies lies on the straight line joining their centres of gravity and divides that line in the inverse ratio of the weights of the bodies. Thus b G bears 29 to i a the same ratio that the weight of A B E bears to that of" ABODE ; i.e., the ratio of w to W. Similarly, & G^ has to 6 c the same- ratio of ?.y to W. Hence 6 G is to 6 a as & G^ is to 6 c, and there- fore G Gi must he parallel to a c. Again, from the similar- triangles 6 G Gi and 6 a c, G G^ is to a c as 6 G is to & a, or as w is- to W. Expressing this mathematically, ^A=^. Whence GG,=:^xac=^feet. a c W * W W Applying the foregoing principle to the case of a ship, it is seem that if any weights are moved vertically, the centre of gravity of: the whole ship will move in the same direction, and will therefore still remain in the vertical line through the centre of buoyancy.. The vessel will thus continue to float at the same water line as; before and with the position of the metacentre unchanged, but with an increased or diminished metacentric height according a» the weights are moved downward or upward. The lowering or raising of masts and yards is an illustration of this, the conse- quent alteration in the position of G being easily calculable by the above formula when the weight moved and the distanca- through which it is transferred are known. If weights are moved transversely across a ship, her centre of gravity will travel out of the vertical line through the centre of buoyancy for the upright position, and the vessel will thus heel over to one side. In Fig. 15 let G and B denote respectively the- centres of gravity and buoyancy of a ship when upright, and. f/C./S. FlG.16. a suppose she is held in that position whilst a weight of w isms- is moved transversely from a to &, a distance of h feet. The- centre of gravity of the vessel, after the weight is shifted, will be- at Gi where G G^ is parallel to a 6 and equal to ^"^^ ,W being- 30 tlie total weight of the ship. As the buoyancy and weight are noA>?. -acting in different vertical lines, the ship, on being released from constraint, will incline until the centre of buoyancy is brought Tertically beneath Gj by the altered shape of the displacement. Let Fig. 16 show its position when this is the case, B^ being the new centre of buoyancy. If the weight moved is moderate in amount,, the consequent inclination is small, and the new line of action of the buoyancy will therefore pass through the metacentre, marked M in the Figure. It is now easy to estimate the angle to which a iship of known displacement and metacentric height will be inclined by moving given weights transversely. For from Fig. 16 P P it is seen that the tangent of the required angle is t:^-^, so that, calling the angle of inclination 6, and substituting for G G^ its value as given above, , a G G, wxh tan. a = „ A = ^ GM "WxGM" To take an example, suppose that in a ship for which W = 9,000 tons and G M = Sj feet, a weight of 60 tons is moved transversely through 28 feet. Here , ^ 60x28 4 ^„ ' *^^-^ = 9000^=75 = '^^^- Referring to a table giving the value of the tangent of each :angle, it is found that the angle of which the tangent is "053 is rather more than 3 degrees, which is therefore the angle required. On the other hand, if the angle to which the movement of a known weight inclines a ship of given displacement — but the metacentric height of which is not known — is observed, the value of G M may be calculated. For transposing the above equation, GM= ^^ ^ ^' feet. W X tan. d For example, moving 60 tons 50 feet across the deck of a ship of ■^,000 tons displacement, causes her to incline to an angle the tangent of which is -^ (about Z\ degrees) : what is her metacentric height ? HereGM= fQ^^Q = 6 feet. 8000 X ^-g This is .the principle of what is known as the inclining .experiment, which is undertaken when a ship approaches com- pletion, in order to ascertain the exact position of the centre of ..gravity in that condition. Then, knowing what weights have to 31 go on board, and the weight of ballast &c. to be removed to com- plete the ship, the centre of gravity in the sea-going condition can be calculated by the aid of principles already mentioned (see page 25), and the metacentric height in that condition seen to be sufficient or not. When a ship is designed, a very detailed calculation is made to find out the final position of the centre of gravity ; but as during construction several deviations are usually made from the design, it is considered advisable to check the result by a direct experiment. In this experiment, a large number of pigs of iron ballast are arranged on the deck in two equal batches, one on each side of the ship, so that the vessel is still upright when this ballast is on board. The ship is then inclined by moving the ballast on one side to the other, the inclination being measured by means of plumb lines suspended in hatchways or other suitable places. FlG. 17. Pig. 17 shows one such line, hung from a batten B, with a second horizontal batten C placed vertically beneath B on which the deviations of the line may be marked. If on moving certain weights from one side of the ship to the other, the plumb line travels from its position a b in the upright to a c, the tangent of the angle to which the vessel is inclined. be is evidently — r- The distance b c ab \e being measured, and that between the battens (a b) being known, the re- quired tangent can be calculated. Usually three plumb lines are erected in the ship to ensure accuracy ; one being forward, the second amidships, and the third aft. Before the ship is inclined, several precautions have to be taken, the chief of which are to see that there are no weights free to m.ove as the ship inclines, such as water in the bilges or boilers, or hanging weights as rigging, boats, &c. ; to select a calm day for the experiment, and to see that the ship is so disposed, as to be affected as little as possible by any light wind that blows. These precautions having been observed, and the ballast and plumb lines being on board, the following is practically the procedure iii any particular ship. The drafts forward and aft are carefully noted, and the mean draft thus obtained. Thence the weight (W) of the vessel is known for the experimental condition, from the curve of 32 displacement. Also the F1C.I8. position of ihe metacentre can be found for this con- dition from the meta- centric diagram, assuming that the trim is not very- different from its value in the designed load con- dition. Let Fig. 18 sho-w the metacentric diagram of the ship, W L being the water line at the time of the experiment. Then M marks the height of the metacentre for the experimental condition. The ballast (of weight w) from^ one side of the ship is now moved to the other through a certaii?, distance ill feet say), and the tangent of the corresponding angle^ of inclination (tan &) is ascertained by means of the plumb lines. Then the metacentric height of the vessel at the time of the- experiment is given by w Y. h GM=. -feet. "W X tan. 9 If now this distance is set down below M in Fig 18, the point Gr so found will be the vertical position of the centre of gravity of the ship in the experimental condition, and thence, knowing the weights to be put on board and taken off to complete the ship, the- position Gj of the centre of gravity in the fully equipped condition can be calculated. In practice, after the deflection due to moving- the ballast which is on one side to the other is noted, this ballast is restored to its original position, when the ship should regain the- upright if no weights have moved during inclination. The ballast on the opposite side is then traversed across the ship through the same distance as the first ; and if this gives a slightly difCerent angle from the first experiment, the inean of the two results is- taken to substitute for tan. 8 in the above equation. It should be noticed that if the trim of the ship, when the experiment is performed, is greatly different from that in the designed load condition, independent calculations must be made for the displacement and position of the metacentre, instead of obtaining those quantities from the curve of displacement and metacentres respectively. A comparison between the positions of the centre of gravity as obtained by calculation and experiment, shows that they usually agree within very narrow limits. For example, in the Devastatiov the difference was 1^ inches, and in the Alexandra 1 inch. 33 fl(i.J9. Lfeet *! Finally, the position of the centre of gravity of a ship is altered by any movement of weights longitudinally, causing the vessel to change trim. Suppose a weight of w tons originally at a in Fig. 19 is moved forward to h through a distance of h feet, the total dis- placement of the ship being W tons. As before, the centre of gravity of the ship will move from its original position G, parallel to ah, through a distance G G^ = ^ ^ — feet. The ship will, in consequence, change trim until the centre of buoyancy moves from its original position B to B^ vertically beneath G^. Let W^ Lj be the water line when this is the case, the water line before changing trim beiiig shown by W^ L^. The vertical through B^ will intersect the originally vertical line through B in some point M^ which is called the longitudinal metacentre, and the distance between G and Mj is known as the longitudinal metacentric height. This is very much greater than the G M for transverse inclinations, being approximately equal to the length of the ship, in a vessel of ordinary proportions ; it is rather less than that length in a short ship of deep draft, and considerably more in a long ship of light draft. As a matter of fact the metacentric height of a ship is greatest for longitudinal directions of inclination, and least for transverse directions, and hence the reason for devoting attention mainly to the latter. As the weight w is moved forward, the ship will be depressed forward, and will rise S 11276 34 aft so that the same total displacement may be maintained. Hence the two water lines W^ L^ and W^ Lj will intersect'' in a point S, this being the centre of gravity of the water plane W^ L^ when the inclination is small, and is usually situated a few feet abaft the middle of the length. For all but extreme changes of trim, however, no great error is involved by assuming that the water lines cut at the middle of the ship's length, in which case the rise of the ship aft (W^ d) is equal to the depression (L^ Lj) forward. Of course if the weight had been moved aft the depres- sion would have been aft and the rise forward. The trim being the difference between the drafts forward and aft, it follows that the total change of trim will be the sum of the increase (Li Lj) of draft forward, and the decrease {W<^d) aft. Let x feet be the total change of trim, and L feet the length of the ship. Then if Wj C is drawn parallel to W^ L^, L2 G is equal to x, the total change of trim, and ^ is the tangent of the angle to which the ship is inclined. L But ^ is also the tangent of the same angle. TT ^_Gi-Gi_ wxh Hence—- g^ "WxttMi and therefore the change of trim (x) is =^2i-^ feet. W X G Ml ^ ^ If the longitudinal metacentric height for the particular ship is not known, a fair estimate of the change of trim may be obtained by remembering that G Mj is, roughly speaking, equal to L, and thus approximate change of trim = = inches. The following example will illustrate the use of these rules : — A ship of 8,000 tons displacement, 300 feet long, and having a longitudinal metacentric height of 315 feet, has a draft of 24 feet forward and 26 feet aft. If now a 45-ton gun is moved aft through 110 feet, what will be the new draft forward and aft ? Here change of trim = ^^ ^ t^n ^^S,k ^^^ = 7 inches approx. 8000 X 315 That is to say, the ship will rise 3^ inches forward and sink 3^ inches aft. Thus, final draft ft. ins. ins. ft. ins. Forward = 24 - 3^ = 23 8^ Aft = 26 + 3^ = 26 3i Had the vessel in Fig. 19 been forcibly held at Wi Li until the weight was placed at &, the moment of the mechanical couple then 35 formed by the "weight and buoyancy would have been W x GGi, or its equivalent wxh,- this is called the "moment to change trim," and its necessary value to produce a given change of trim can be estimated from equation (I). For, transposing that equation, W X G Mr the moment to change trim (w x h) is = — ^a f~^ ^ ^ and thus can be calculated when a; is known. To find the moment to change trim one inch, put x = 1. Then W X GMi moment to change trim one inch is =-^r?; 5^ foot tons. 12 X L From this it is seen that ascertaining from equation (I) the change of trim in inches due to moving certain weights through known distances, consists in dividing the moment to change trim (w X h) by that necessary to change trim one inch. Remembering the approach to equality of the length of war ships of usual proportions and their longitudinal metacentric height, the moment necessary to change trim one inch becomes W approximately = =-^ This may be of use for rapidly deter- mining whether any possible movement of weights on board a particular ship, of which only the displacement is known, will enable her to get over an otherwise impassable bar at the mouth of a river or harbour. For instance, suppose the ship mentioned in the preceding example can only pass over a particular bar when drawing not more than 25^ feet of water aft, and that it is possible to move the following weights the given distances forward : — 20 tons, 200 feet ; 25 tons, 100 feet ; 15 tons, 70 feet ; and 30 tons through 50 feet. Will the movement of these weights be sufficient ? Since the ship has to rise 6 inches aft, there will be an equal alteration forward, so that the change of trim must be 12 inches. Now, the moment necessary to make this change is approxi- W mately = JI x 12 = 8000 foot tons. And the moment to change trim provided by moving the weights is 20 X 200 + 25 X 100 + 15 x 70 + 30 x 50 = 9050 foot tons. Hence it may be concluded that by moving these weights the vessel may be taken over the bar. Of course, if the vessel is very long and of light draft, an allowance must be made for the differ- ence between its length and the longitudinal metacentric height, as previously mentioned ; this necessitates, moving more weights on board than would be requisite under ordinary circumstances. It may be remarked in passing, that the determination of the trim given to a ship by putting weights of moderate amount on S 11276 C 2 36 board, may be reduced to the simpler case of moving weights already on board as follows : — The weights are supposed first placed in the same vertical line as the centre of gravity of the water plane area, in which position they will sink the ship deeper in the water without change of trim. This sinkage can be readily estimated when the tons per inch at the water line is known. Then the weights are supposed distributed to their proper places in the ship, and the consequent moment to change trim calculated in a manner similar to that adopted in the preceding example ; observing that if some weights are moved forward and some aft, the difference of the moments forward and aft must be taken as the final moment to change trim. To take an example : — On a ship which has a draft of 25 feet forward and 27 feet aft, there are weights of 70, 90, and 20 tons to be placed. The first is to go 30 feet before the centre of gravity of the water plane area, and the second and third are to go 50 feet and 40 feet respectively abaft that centre of gravity. If the tons per inch at the load line is 40, and the moment to change trim one inch = 800, what will be the draft of water forward and aft, after the addition of the weights ? The total weight put on board is 180 tons, and therefore, sup- posing it so placed as to cause no change of trim, 180 Sinkage =— — = 4^ inches, and the draft would then be : Forward, 25 ft. 4^ ins., and Aft, 27 ft. 41 ins. If now the weights are supposed removed to their proper places The moment tending to trim the ship') „„ by the stern is - - - ' J90 x 50 + 20 x 40 = 5300 The moment tending to trim the ship > _„ or. ,,,,,. ^ S 70 X 30 = 2100 by the head is - - - -3 /. Resultant moment trimming the ship by the stern = 3200 Whence change of trim = ^^ = 4 inches. The final drafts required are, therefore, ft. ins. ins. ft. ins. Forward - 25 4^ — 2 = 25 2^ Aft - - 27 4^ + 2 = 27 6i Reverting now to Fig. 10, it will be remembered that when the ship is at any angle, the couple W x GZ is the statical stability at that angle. The weight of the ship being known, this stability can be calculated when the length of the arm G Z is known. For small angles, where the line of action of the buoyancy passes i1 through the metacentre, G Z is equal to G M sin. 0, so that for angles to which this " metacentric method " applies, the statical stability is W X G M sin. 9 foot tons. For larger angles than this, the lines of action of the buoyancy will not pass through the metacentre, and then to determine the values of G Z a difiEerent method must be pursued. Fig. 20. In Fig. 20, suppose Wi Li to be the water line at which a ship floats when upright, B marking the position of the centre of buoyancy for that condition. Also suppose that after the vessel is inclined through a considerable angle, d, the water line becomes Wj Lj, the displacement remaining the same as before. Since these displacements are equal, the remainders after taking away the volume W^ S Lj K Wj, which is common to both, must be equal. That is, the wedge-shaped volume Wi S "Wj (known as the wedge of emersion) must be equal in volume to the wedge Li S Lj (the wedge of imtnersion), although of a different shape. Let g^ be the centre of gravity of the wedge of emersion and g2 that of the wedge of immersion, and join giffn. It is seen that the dis- placement up to "Wj Lj differs from that up to W^ Lj only in having the volume of the wedge Wj SWj transferred from g^^ to g^. Hence, if V is the volume of displacement, and v that of each wedge, the centre of buoyancy of the ship in the inclined position will be — by the principle already established — at B^, where B Bj^ is drawn parallel to g^ g^, and is of a length given by BBi = ?i^ 38 The line drawn through Bj perpendicular to W^ Lj Avill be the new line of action of the buoyancy, and G Z, drawn at right angles to this from the centre of gravity, G, will be the required arm of the righting couple. To obtain the length of this arm, draw g^ h^ and g^ \ perpendicular to W^ 1,^, and B R parallel to G Z. The distance \ \ is the horizontal distance through which the wedge is moved, and B R is the transfer of the centre of buoyancy in a horizontal direction. Hence, by an extension of the principle mentioned before, the length of B R is V X W V So that if C is the point of intersection of the line of action of the weight with B R, GZ=CR=BR-BC =iiiiAA_BGsin. 61 For any particular ship the positions of B and G are known, as well as the displacement, so that B G and V are known quantities in the above equation. Also, for any given angle of inclination sin. d is known, and thus the determination of the value of G Z for any angle resolves itself into finding the value of the product V X hj^h^ for that angle, a proceeding involving rather laborious calculation. The length of G Z obtained as above varies, of course, according to the inclination ; but as, for equal angles on opposite sides of the vertical, the values of G Z are obviously the same, those values need only be investigated for inclinations on one side. It should also be stated that when the angle through which a ship is inclined is considerable, the line of action of the buoyancy may fall to the left of G, Fig. 20, in which case the stability is said to be negative, since the weight and buoyancy form an upsetting instead of a righting couple. The lengths of G Z are calculated for every 8 or 10 degrees of heel, until the stability becomes negative, and the results are fhown on a curve of statical stability as shown on Plate I., Fig. 21. This is constructed in the following manner. A. horizontal base line A B is taken, on which to set off angles of inclination, measuring from the point A, the scale employed being usually such that a quarter of an inch represents one degree. Let A C denote, on this scale, one of the angles for which the length of arm of righting lever has been found. From C draw D at right angles to A B, and of such a length as to indicate to scale — a quarter of an inch usually representing one-tenth of a foot — the length of FlaU I. J^.2Z, M^. ?S. Fi^. 24-. AjtCfLe* i pF In/cUrvcvtiATn^ J& Fi^.25. Fl^.2^. i^ 2h^X<3^ /»a^ 08. JuddlCLilK 7Ji75 Fai-nrgdon R'' i Doctors Commons 3^S4-.3.9f, 3d Gr Z at the given angle. If, as is assumed, the couple formed hy the weight and buoyancy at the angle denoted by A C tends to turn tht) ship towards the upright position, C D is set above the base A B as in Fig. 21 ; had the stability been negative at that angle, C D would have been set helow A B. A similar construction being made for each of the other calculated values of G Z, a curve A D F H K L drawn through all the points such as D, is the required curve of statical stability, or, shortly, the curve of stability. Fig. 21 shows the curve for the vessel indicated in Fig. 22, the upright position being supposed one of stable equilibrium. This exhibits the usual characteristics of such curves, and the varying position of the line of action of the buoyancy with regard to the centre of gravity can be traced by its means. In the upright position there is no arm of righting couple, the curve of stability meeting the base line at A. As the vessel is inclined, the line of action of the buoyancy falls to the right of G, so that the weight and buoyancy form a righting couple, indicated by the curve rising at first above A B. The value of G Z rapidly increases up to the angle denoted by A E, this being the angle at which the deck edge becomes immersed as depicted in Fig. 23. Beyond this, the increase of G Z is less rapid, and finally ceases, G Z attaining a maximum value at the angle represented by A G, where the tangent (a b) to the curve is horizontal. At this angle, therefore, the effort of the ship to right herself is greatest. As the inclina- tion is further continued, the value of G Z diminishes,, and the line of action of the buoyancy will eventually pass through G again as in Fig 25, in which case the vessel will be in equilibrium once more, the value of G Z being zero. Finally, beyond this second position of rest, the buoyancy will fall to the left of G, and will operate, together with the weight, to heel the ship farther from the upright position. This fact is indicated on the curve by its being below A B. Hence, if the ship is inclined by external forces to any angle less than that represented by A K, she will return to the upright when the external forces cease to operate. If, however, those forces incline the ship beyond that angle, she will overturn. The angle A K is known as the angle of vanishing stability, and its amount is the range of stability. This is one important criterion as to the probable safety of the vessel at sea. Other elements of the curve influencing the ship's safety are the maxi- mum length of the arm of righting couple and the angle at which this occurs. Other things being equal, the greater the range of stability and maximum value of G Z — in other words, the greater the area of the curve of stability — the greater is the pro- 40 bability of the vessel's safety at sea. This area is therefore made as large as possible consistently with other conflicting conditions. For example, since for the first few degrees from the upright, the value of G Z is very approximately equal to G M sin. 6, it is seen that the greater the metacentric height is made in a ship, the steeper will the front part of the curve be, and the greater its area. But the adoption of a great metacentric height leads to heavy, rolling at sea. So, in adjusting these clashing interests, the experience gained by successful ships is mainly relied upon ; remembering the general principles that a mastless ship requires less stability than one carrying a large spread of canvas, and one with large resistance to rolling — due to bilge keels, &c. — less than one in which this resistance is small (see Chap. III). It was remarked that a ship is in equilibrium when inclined at her angle of vanishing stability ; that equilibrium is, however, evidently unstable, since, if the vessel is slightly moved in either direction from that p.„ ^y position she will . return to the up- right, or overturn, according to the direction of incli- nation. It thus appears that whilst the points in which the curve of sta- bility cuts the base line denote angles at which the ship will be in equilibrium, the character of that equilibrium is indicated by the relative positions of the curve and base line. If in passing towards the right frona. a position of rest the curve lies ahove the base, the equilibrium is stable ; if helow it, unstable. A class of vessel for which a curve of stability can be easily constructed is that designed for submarine service. Let Fig. 27 show a transverse section of such a one, having its centre of buoyancy for the upright position at B, and its centre of gravity at G. If the vessel is inclined, the shape of the displacement is not altered in the least, and therefore the centre of buoyancy for the displaced position is^also at B, through which the new direction of the buoyancy acts. It is thus seen that, in a sub- marine vessel, the metacentre coincides with the centre of buoyancy, so that for stable equilibrium in the upright, the centre of gravity must fall helow B. Again, since the buoyancy acts 41 through B whatever the inclination, the arm of righting lever for any angle 6 is equal to B G sin. 6; and as B G is known for any particular ship, and the value of sin. 6 can be obtained from mathe- matical tables, the lengths of G Z can be readily calculated. The stability will be a maximum when sin. 9 is greatest, and will vanish when sin. 6 is zero. The former is the case when ^ = 90 degrees, sin. 6 being then equal to 1 ; and the latter when = 180 degrees (G being then vertically above B) as well, as when the ship is upright. The equilibrium when 6 = 180 degrees is evidently unstable, and on the slightest disturbance from that position the vessel will return to the upright. Such a vessel, then, has its maximum stability at 90 degrees, a range of stability of 180 degrees, and possesses the property of being self; righting, that is, of returning to the upright from any other position in which she happens to be. In passing, it may be: mentioned that lifeboats have this quality, of gelf-righting, the air spaces fitted along their sides and the considerable sheer (or curve upwards) given to their ends bringing the metacentre so low down, when the boat is keel upwards, that it falls below the centre of gravity, making that position one of unstable eqtdlibrium. In obtaining the curve of stability Of an ordinary vessel, of the parts lying above the water only those which can be made water- tight — ^that is, the parts contributing the reserve of buoyancy — are reckoned as assisting to give stability to the ship ; and the following further assumptions are made. The centre of gravity of the ship is supposed to remain in the same position throughout, which assumes that no weights fetch away as the ship is inclined ; and it is also taken for granted that no water enters into any watertight space either above or below the water line, through ports, scuttles, hatchways, &c., as the vessel is heeled over. On the first of these two latter assumptions Mr. White (Manual of Naval Architecture, page 123) remarks : — " This may be considered an improper supposition, especially in cases where stability is maintained beyond' the inclination of 90 degrees from the upright ; but it is to be observed that such extreme inclinations are not likely to be reached, whereas for less inclinations the supposition affects all classes similarly." And as to the second, "This assumption' is fair enough as regards most of the openings, which are furnished with watertight covers, plugs, &c. ; and as regards some of the hatchways which are usually kept open even in a seaway, it is only necessary to remark that they might be battened down on an emergency, while their situation near the middle line of the 42 deck prevents the water from reaching them except at very large angles of inclination." (pp. 123-4). On these assumptions the curves shown in Fig. 28 for four of H.M. ships and the American ironclad monitor Miantonomoh have been obtained. A considera- Fic.za. ao 90 tea 30 40 so so 70 ANGL£S OF INCLINATION References. 1./ Captain. 3. Inconstant. 5. Monarch. 2, Glatton, i. Miantonomoh. tion of these curves will bear out the statement previously made that the steepness of the front part is dependent upon the metacentric height. That for the Miantonomoh is steepest, this vessel having a metacentric height of about 14 feet. Of the others the Captain had a height of 2"6 feet, and the corresponding information for the remainder has been previously given (page 23). A comparison of the curves for the Captain and Monarch exhibits the great influence of freeboard or height from the water line of the upper deck amidships at the side, on the area and range of curves of stability. The displacements of these two ships only differ by 4 or 5 per cent., and their proportions are not greatly different, the great variation in the two designs being in the matter of freeboard. The Monarch has 14 feet, whereas the Captain possessed &\ feet only. In consequence of the slightly smaller metacentric height of the former, her curve of stability falls at first below that for the Captain ; but she has a total range of stability of 69^ degrees, reaching her position of maximum stability at 40 degrees — where the G Z is 21f inches — whereas the Captain has a total range of 54^ degrees only, and reaches her position of maximum stability at 21 degrees, at which angle she has a G Z of only lOf inches. The reason for this influence of freeboard is not difficult to understand if reference is made to Fig. 21. For the greater the freeboard, the greater will be the angle at which the deck edge becomes immersed and I/O 43 the farther will the portion A D F of the curve be prolonged before that curve begins to turn round towards the base as at F. Of the other vessels, the curves of which a,re given above, the Inconstant has a freeboard of 15^ feet, the Miantonomoh 3 feet, and the Glatton also 3 feet, although this last ship has in addition a superstructure on the upper deck amidships not extending to the side, a fact which accounts for the irregularity in the curve of that vessel. It must also be remarked that in comparing ships of greatly different proportions, the freeboard' of each ship must be viewed with reference to its breadth ; it being evident that 10 feet of freeboard on a ship 50 feet broad will be more influential than the same amount in one 70 feet in breadth, on account of the larger angle to which the former must be inclined before the deck edge becomes immersed. The freeboard must also be compared with the total depth, a given amount having a greater relative effect in a shallow than in a deep ship. The comparison instituted between the Monarch and Captain serves to indicate the great influence which reserve of buoyancy — of which freeboard is to some extent a measures-has upon the curve of stability, but one or two further illustrations may not be without value. For example, in vessels of the Canada class, if the reserve of buoyancy is increased by making the poop and forecastle watertight, the range of stability is increased by about 16 degrees from its value when those erections are open ; and the area of the curve is also enlarged. Again, the curve marked A in Fig. 29 is the curve of stability for a vessel having a superstructure extending from side to side above the upper deck for rather more than one-third its length — a vessel somewhat similar, in fact, to the one shown in Fig. 51. If the superstructure is thoroughly destroyed by shell fire, the material falling upon the upper deck FlO.29. R-2-ff o lo to 3o «^7 so eo 7o will cause the centre of gravity of the ship to move downwards, and will thus increase the metacentric height. At first, therefore, the stability of the damaged ship compares favourably with that for the intact condition, but in consequence of the decrease in the reserve of buoyancy, the area and range of the curve of stability is lessened to the extent shown by the curve B. Further, if the reserve of buoyancy could be totally destroyed in a vessel, a simple consideration will show that she must inevitably capsize. For, imagine a complete watertight deck to be worked at the height of the water line, and the part above this demolished. The vessel would still continue to float, since the buoyancy remains uninjured, but, as she is now in the condition of a submarine vessel with the watertight deck awash, her metacentre will be at the centre of buoyancy, the position of which latter point has, of course, not been altered by the damage above water. Now in ordinary vessels the centre of buoyancy is several feet below the centre of gravity, and therefore the ship under consideration is in unstable equi- librium when upright, so that, like other submarine vessels in similar circumstances, she must capsize. The hypothetical deck mentioned above very closely resembles the protective deck fitted in many ships ; and although of course it would be practically impossible for a vessel to become reduced to the above condition, looking at the matter in this light enables the importance of the reserve of buoyancy to be better recognized, and shows the great necessity for protecting against too extensive damage that part of the ship lying above water. This protection may be given by armouring the sides ; by minutely subdividing the space above water so that any damage may be localised ; by stowing stores in that space, in order that, even after it may be laid open to the sea, a reserve of buoyancy equal to the volume of the stores may still be retained ; or, finally, by any combination of the others. These arrangements will be noticed in Chapter XIV. Curves of stability are not only constructed for the fully laden condition, but also for the extreme light condition of the ship. "When the stores are consumed the vessel has additional freeboard due to her rising in the water ; this fact tends to give to the curve for the light condition greater area and range than that for the load condition. On the other hand, the metacentric height is usually less for the light condition, leading to a reduction in the size of the curve for that condition. Of these two opposing tendencies, the latter is the more potent, generally speaking, the curve for the light condition having less area and range than that for the load condition, although the reduction is not great. It has before been remarked that some merchant ships in the light condition are unstable when upright. Such vessels will have a curve of stability for that condition similar to the one shown in 45 Fig. 30, the ppirit A, where the curve begins to rise above the base line, denoting the angle to which the ship will loll to one side or Fig. 30. other of the vertical before attaining a position of stable equi- libriTim. After passing this angle, the vessel will usually have a good range of stability, so that it can by no means be inferred that an ordinary ship can be easily capsized if her upright position happens to be one of unstable equilibrium. Attention will next be directed to what is knovm as dynamical stability. The amount of mechanical work necessary to heel a ship over to any angle from the upright position, is called the dynamical stability at that angle. Work is said to be done when a resistance is overcome through space, and is usually expressed in foot-tons when dealing with large amounts. For instance^ in raising a weight of one ton through a vertical distance of one foot, one foot-ton of work is performed ; if through two feet, two foot- tons, and so on, the amount of work always being estimated by multiplying the resistance by the distance through which it has been overcome. The same term " foot-ton " has been used with a different meaning when speaking of statical stability, and the nature of the difference may be illustrated by a simple example. In Fig. 31 let A and B be the ends of two capstan bars, at each of which a man exerts a pressure P in the direc- tions shown, when just balancing the force acting upon the body of the cap- stan. The product PxAB measures the effect of the couple exerted by the men in foot-tons, if P is expressed in tons and A B in feet, and corre- sponds to the expression (W X G Z) for the statical stability of a ship. So long, however, as the men stand still at A B no resistance is overcome Fig. 31. 46 through space, and hence no work is performed. Suppose now the men, each steadily exerting the force P, move the bars from A B to Aj Bj. In the new position, the magnitude of the couple exerted by the men is still P x A B foot-tons, but the worlc done in foot-tons in going to that position differs altogether from this. The amount accomplished by the man at A is P multiplied by the length of the circular arc A Aj, and by the one at B is P multiplied by the length of the arc B B^. The length of the arc A Aj is equal to A C multiplied by the circular measure of the angle A C Aj, and B Bj is obtained by multiplying B C by the circular measure of the same angle. Hence Total work done = P x (A Aj -h B Bj) foot-tons = Px(AC4-CB)x circular measure of the angle A C A^ = (P X A B) X circular measure of the angle A C A^. It is thus Been that the work done in foot-tons by a couple turning through a certain angle is equal to the moment of .the couple in foot-tons multiplied by the circular measure of that angle. In the case of a ship, if fluid resistance be neglected, the work done in inclining her to any angle is expended in overcoming the resistance offered at each instant by the statical stability, and differs from the preceding example only in the fact that the moment to be overcome varies from instant to instant in this case, owing to the variation in the length of the arm of righting couple. To investigate the amount of work done during the inclination of a ship, a modification of the method adopted in the above example is resorted to, by which means it may be shown that the area of the curve of statical stability up to any angle, measures the dynamical stability at that angle. For example, if A B C, Fig. 32, is the curve of statical stability for a ship, the area of the part A a B, when properly interpreted, will give the dynamical stability for the angle A a. If the distance ab is set up to represent on a certain scale the area AaB, and a similar construction followed out for several other angles, the curve A ft D drawn through all such points as b, is the curve of dynamical stability, which gives for any angle the dynamical stability at that angle by simple measurement. From the foregoing it is seen that the greater the area of the curve of statical stability, the greater will be the amount of work necessary to be done on the ship to incline her to the angle of vanishing 4=7 Ft 0.32. stability, and the less likely therefore is she to be inclined to that critical angle by the action of external forces. Thus is demonstrated the importance of the area of the statical stability curve in relation to the safety of the vessel. Brief reference will now be made to the effect on the stability of damage to a ship at or below the water line. If the injury occurs above a watertight fiat, so that the surface of the water is at the same height inside as outside the ship, the area of the intact water plane is thereby decreased, and the distance between the centre of buoyancy and metacentre thus shortened (see page 21) ; on the other hand, as the ship sinks in the water, consequent upon the damage, the centre of buoyancy rises, and thus the final position of the metacentre and value of the metacentric height (which determines the stability at small angles) is depen- dent upon the relative effect of loss of water plane area and rise of centre of buoyancy. As an extreme example, reference m^ay be made to the Inflexible. If her unarmoured ends are com- pletely riddled at the water line, her metacentric height falls from 8J to 2 feet, indicating the importance of protecting the water plane area from, excessive damage, and of providing vessels of the same type as Inflexible with greater metacentric heights than would otherwise be necessary. If, however, the damage is low down, below a watertight flat, so that the compartment is completely filled, it will make no difference to the stability if the damaged part is afterwards made good. Conceive this to be the case, the admitted water being virtually so much ballast low down in the ship. This will therefore considerably lower her centre of gravity, whilst the sinkage due to the admission of the water will but slightly affect the position of the metacentre. Such injury will thus increase the metacentric height, and, although the freeboard has 48 been somewhat diminished, the area and range of the curve of stability will be little if at all diminished from their values when the vessel is intact. If water in the hold of a vessel is free to shift, the stability at any angle is less than it would have been had the water been fixed in position, because, in travelling towards the side to which the vessel is inclined, the water carries the centre of gravity of the whole ship nearer the line of action of the buoyancy. Examples. I. The water lines of a vessel are 3 feet apart, and the depths below the respective lines of the corresponding centres of buoyancy are 10*5, 9'0, 7-5 and 6-1 feet, beginning with the load water line. The distances between these centres of buoyancy and their corresponding metacentres are, respectively, 12'3, 13'9, 16"0, and 19 feet. Construct the metacentric diagram to scale. II. An inclining experiment, performed on the ship of the preceding question when floating at a line 4 feet below her load water line, showed that in this light condition, a weight of 31 tons moved 40 feet across the vessel from side to side inclined her to an angle the tangent of which was -^. The weight of the ship at the time of the experiment (including inclining ballast) was 5,952 tons ; what was her metacentric height at that time ? Again, if in putting the necessary weights on board to sink the vessel to her load water line from the light condition, her centre of gravity is lowered 6 inches from its position in that condition, what is the metacentric height of the vessel when fully laden ? Answers — 2"5 feet ; 2'8 feet. III. A weight of 68 tons, already on board a vessel 315 feet long and of 7,497 tons displacement, is moved forward through a distance of 105 feet. What change of trim is thereby produced, the longitudinal metacentric height being 300 feet ? Answer — 12 inches. IV. If before moving the weight in question III. the draft of water had been 23^ feet forward and 25^ feet aft, what would be the new drafts after that weight was moved ? Answer — 24 feet forward ; 25 feet aft. V. The weights enumerated are to be placed on board a ship drawing 25 feet of water forward and 26 feet aft, in the positions 49 specified relatively to the centre of gravity of the load water plane ; viz.,. 30 tons 20 feet before and 40 tons 30 feet before ; also 50 tons 70 feet abaft and 20 tons 90 feet abaft. Assuming the tons per inch to be 35, and the moment to change trim one inch 700, what will be the new draft of water forward and aft ? Answer — 25 feet 1^ inches forward ; 26 feet 6^ inches aft. VI. Construct to scale the curve of stability of which the ordinates at intervals of 10 degrees are, starting from the upright, 0, 1-09, 2-3, 2-98, 2-72, 1-84, -57 and — -89 feet respectively. Also, find (1) the range of stability and (2) the maximum length of arm of righting couple and the angle at which it occurs. Answers — 64 degrees ; 3 feet ; 32 degrees nearly. S U276 50 CHAPTER III. OSCILLATIONS OP SHIPS. The principal oscillations of sMps take place in the transverse and fore-and-aft directions ; in the former case the motion is termed rolling, and in the latter x-iitching or 'scending according as the bow of the ship moves downward and the stern upward, or vice versa. Of these oscillations rolling is much the more impor- tant, and will alone be dealt with here. The extent of the rolling motion is measured by the inclinations which the ship reaches from the upright, and in going from her extreme position on one side of the vertical to the next extreme position on the opposite side, she is said to perform a single oscillation. The arc of oscillation is the total angle swept through by the ship in an oscillation, and her period is the time occupied in performing one oscillation. For example, the Devastation was found to move from an angle of 10 degrees on the starboard side to 8 degrees on the port side in 6-76 seconds ; her arc of oscillation at that time was therefore 18 degrees and her period G'76 seconds. The main object in studying this branch of NaAal Architecture is to ascertain what elements govern or affect the rolling motions of a vessel in a seaway, so that her probable safety at sea may be secured. But in approaching this question, although it is evident at once that a ship will experience resistance when rolling through the water, it is convenient to consider first the purely hypothetical case of a vessel rolling unresistedly in still water, where the only two forces acting upon her are the equal ones due to her weight and buoyancy ; secondly to enquire, by means of experiments on actual ships or models in still water, in what respects the conclu- sions arrived at in the above imaginary case are modified by the operation of resistance ; and lastly to see in what manner the oscillations are afEected by the presence of waves. This order will therefore be preserved in the following remarks. First then, consider a ship rolling in still water subject only to the two forces of the weight and buoyancy. Suppose her forcibly inclined to the position shown in Fig. 34. To do so requires the Fig. 3* Fig. 36. expenditure of worh or energy, the amount of which can be ascer- tained from the curve of dynamical stability, and ■which could be again recovered from the vessel if she were attached to, say, suit- able pulleys and thus forced to raise a weight when released from her inclined position. She has, in fact, acquired energy hy virtue of her position from the vertical. Now by a well known principle — that of conservation of energy — if no external resistances act upon a body, the energy due to position added to that due to motion remains a constant quantity. Applying thi^ to the case in question, suppose the unresisted vessel suddenly released from the position Fig. 34. At that instant there is no motion, and conse- quently her total energy is the dynamical stability in that position. Reaching the upright. Fig. 35, she has no energy by virtue of position, all the dynamical stability having been converted into energy of motion. Finally, on coming to rest on the opposite side of the vertical, Fig. 36, all the energy she possesses is that due to position, so that the dynamical stability there must be equal to that she had when first released. That is to say, the ship must reach, on the opposite side of the vertical, an angle equal to that from which she first started. If, therefore, a vessel rolled unre- sistedly in still water, she would continue to oscillate from side to side without any diminution in the angle of roll from the perpen- dicular. Again, assuming that the front part of the ship's curve of sta- bility is a straight line (an assumption which is approximately true in most ships up to 10 or 15 degrees), a mathematical investi- gation shows that the ijeriod is the same for large as for small arcs of oscillation ; that is, she will swing through a total arc of, say 20°, in the same time as through one of, say 4°. This fact is usually expressed by saying that the ship is isochronous within the limits of roll mentioned above. If the period be denoted by T, its value in seconds is given by the formula ^2" - - k T = -554^- or T = -554- s/va Where m is the metacentric height, and k the radius of gyration when the ship is oscillating about a longitudinal axis through her S.O. 11276 D 2 52 centre of gravity. To explain the meaning of the term radius of gyration, a definition of moment of inertia will be necessary. Conceive the ship to be made up of a -very large number of small parts or elements ; then the moment of inertia, about a longi- tudinal axis through her centre of gravity, is the sum of the products obtained by multiplying the weight of each element by the square of its perpendicular distance from the axis. For example, if w-^, w^, w^, &c., denote the weights of elements situated at distances of a;^, x^, x^, &c., respectively from the axis, the rnoment of inertia is W-^ X X^ + IV^ X X^ + IV^ X x^ + &c. The total weight of the ship is, of course, the sum of the weights iv^, w^, Wj, &c. ; let this be represented by W, and the moment of inertia by I. Then the radius of gyration (k) is such a length that Wk^ = I. In other words, if the total weight of the ship could be concen- trated at a point distant from the axis a length equal to the radius of gyration, it would have a moment of inertia equal to that which the ship actually possesses about the same axis. To take an example of the above formula, the Cyclops class have a radius of gyration of about 18 feet and a metacentric height of 3"45 feet. Hence their time of performing a single oscillation, if rolling unresistedly, would be given by T = -554 /— = -554 x -^ = 5'36 seconds. V 3-45 l-8b The usefulness of this formula will be noted farther on. For very large angles of roll, the period will be increased from that given by the formula for isochronous rolling, this being analogous to the behaviour of a simple pendulum. If a pendulum is of such a length as to have a period of one second when oscil- lating through small arcs, the period becomes increased to 1-073 seconds when the pendulum oscillates 60° on each side of the vertical, and to 1-762 seconds when swinging through 150° on each side. Secondly : An actual ship set rolling in still water differs from the preceding case in that fluid resistance is in operation, in addition to the moment of stability. This resistance is due : — (1) To the rubbing against the water of the surface of the ship when rolling, ■proda.cmg frictional resistance. (2) To the opposition offered to the passage through the water, in a more or less flat-wise direction, of projections such as the 53 keel and bilge keels, and of the comparatiTely flat parts of the ship at the ends : this is known as direct or head resistance. (3) To the creation of waves as the vessel rolls : these cannot be propagated without the expenditure of energy, which must be supplied to the water by the ship, thus afEecting her motion. The first is analogous to the resistance offered to the passage of a thin board edge-wise through water ; and the second, to that offered by the water to the same board moved flat-wise through it. Before tracing the effects of fluid resistance, the methods em- ployed to set ships rolling in still water will be noticed. The flrst, adopted for small ships, is similar to that employed at Brest in carrying out experiments on the Elorn, a vessel of about 100 tons displacement.' The ship was placed parallel to the quay and at some little distance from it, and then a purchase was led in a horizontal direction from a pair of shear legs erected on board (for want of a mast) to a winch on shore, by which means the vessel was inclined to the required angle and then set free. Large vessels are set rolling through the action of a number of men on board, who run from one side of the deck to the other, their movements being carefully timed with reference to the motion of the ship. The men should always be on the descending side of the vessel, so that their weight may help to incline her further, and to accomplish this, their movements should be as follows. To start the inclina- tion they must stand on one side of the deck and then run towards the middle line, reaching it at the time the maximum inclination of the ship occurs ; they should then run up the deck to the op- posite side, stay there as long as possible and get back again to the middle line by the time the ship reaches her extreme angle on that side of the vertical. Thus the positions of the men and the directions in which they should run in relation to the positions of FiG. 37. A the ship are as shown in Fig. 37, at A, from which it is seen that the men should, where possible, always he running up an incline. By this means large ships may be rolled several degrees from the vertical. 54 If, after a vessel has been set rolling by either the above methods, careful observations are made of the successive angles reached from the vertical, it will be seen that instead of these angles re- maining always the same — as they would do in unresisted rolling — the effect of resistance is to continually diminish them, and finally to bring the ship to rest. The reason for this is apparent, for, when a vessel is held to a certain angle on one side of the vertical, it has at that angle an amount of dynamical stability or energy of position which can be obtained as before from the curve of dynamical sta- bility ; but after she is released a portion of this energy is used up in overcoming the various opposing resistances, so that she will come to rest at such an angle on the opposite side of the vertical as will make the difference between the dynamical stabilities at the beginning and end of the oscillation equal to the energy ab- sorbed by the resistances ; the greater the resistances, therefore, the greater will be the degradation of roll, and the sooner will the vessel be brought to rest. The successive angles reached on opposite sides of the vertical during a rolling experiment are carefully noted, until the inclination becomes reduced to two or three degrees, and the results plotted off on a curve of declining angles, or, as it has until recently been called, a curve of extinction. This FfG.38 t^^i^ latter being also securely fastened to the inner flanges of the transverse frames. Near the keel, -where the width of the athwartship frames precludes the use of Z bars, the longitudinals are made up of plates worked inter costally between the frames, each being secured to the outer bottom and to a continuous piece of plate riveted inside the frames, by short pieces of angle bar, as indicated in Fig. 74. In some recent vessels the Z-bar longitudinals have not been worked, their places being taken by short lengths of angle bar which fit between the transverse frames and project sufBciently far inside them to take hold of two continuous angle bars ; a sketch of one of these longitudinals is given in section in Fig. 68. The above completes the description of the main frames of the ship ; and attention will now be directed to the subordinate but important parts of the framing which stiffen the main frames, viz., the beams and pillars. Beams are usually made of long bars of moulded iron or steel and one is worked from side to side under each deck and platform at every transverse frame, except where these in armour clads are only 2 feet apart, and then they are connected to alternate frames only. Those to the spar and upper decks of the vessels under consideration consist of tee-bulb bars 9" deep, the width of the top flange being 5^" ; whilst the beams under the remaining decks and platforms are of angle bulb section of the following sizes : — Under the main and lower decks, 9" deep and 3^" flange, and beneath the platforms 5" deep with a 2^" flange. Each of these beams has a knee or arm formed at each end, and these are connected to the corresponding transverse frames by two or three rows of rivets as Fic.69. 91 at a and b iu Fig. 69, where is exhibited the riveting of a beam arm to one of the frames above and behind armour respectively. "Where tee-bulb bars are employed, a part of the upper flange must be cut away at the ends as shown in plan c. Fig. 69, so that the beam may fit closely against the frames. To save weight, some of the beam arms are lightened by holes, as shown in Plate III. From that Plate it will also be seen that under the spar deck, where no transverse frames are worked, the beams are united to the side plating through the medium of angle bars. In some ships, under concentrated weights such as - _., turrets, &c., the beams are made of ,_^ strong bars of H, or other similar solid, ' T section ; or they are built up "with plates , . and angle bars, three usual sections being JllL given in Fig. 70. Also the beams under platforms in hold are sometimes made of small Z or angle bars. A " round up " from the sides towards the middle line is usually given to the beams, in order that any water on the decks may flow readily into the gutterway. Those under that part of the upper deck between the barbettes are, however, not thus curved, the beams being kept level for convenience in laying the racers for the guns. By supporting a beam in the middle of its length, its strength is considerably increased, and such support is given to each beam in the ship as far as possible, either by utilising the ordinary structural arrangements or by the use of pillars. An example of the former method may be seen in Fig, 73, where the divisional bulkhead between the engine and boiler rooms on opposite sides of the ship efficiently keeps the beams of the main deck in position at the middle line. Most of the pillars used are made tubular, as that form possesses greater lateral rigidity than solid ones having the same weight of material. They have their heads and heels welded in solid, and formed so as to fit against the beams or plating of the decks between which they are worked. Fig. 71 shows in front and side elevation one of the pillars under the 92 Fic.71. Side Elevation- fRom- Elevation. Fic.72. Front Elevation. Side Elevation. o o O PLAN OF HEEL . HEAD VH ZIVw. WOOD DECK 1 upper deck amidships ; these illustrate the most usual shapes of both the heads ; and heels of pillars, and likewise show their fastenings. When a pillar is to be placed under a beam of H or other similar section,iits head is formed similarly to the heel of the pillar just referred to. In the] vicinity of capstans, the pillars must be made capable of being moved out of the wayswhen the capstan bars are to be used. Elevations of one of these portable pillars are given in Fig. 72, from which it will be seen^that it is made to pivot about a bolt in its upper part,:whilst|.its lower end carries a brass nut with a rounded under surface, working upon a screw thread ; which nut, when the pillar is in place, is turned round by means of a spanner until it bears tightly against a brass casting of corresponding shape fixed to the deck plank. It is only necessary to turn back the nut so as to clear the metal casting, in order to hinge up the pillar whenever required ; or, by removing the bolt securing the head, the pillar can be taken away if this is thought desirable. There are various other devices for securing the heels when the pillars are required to give support to the deck above, one of the most common being to drive them into wedge-shaped shoes fixed to the deck. TlaUm Fig. 73. '3?^^^^'^ piaii^ Fi^. 74, 7i) Kjuc£.pci^e. 92/ JuddlC'Llth. 75t75.Farnn9don R^ & DoctorsCommons 2(34 fZ-3l> 93 The largest pillars employed in the above vessels are 7" in diameter and f " thick ; and the smallest 5" diameter and ^" thick, 2. Of a Sheathed Corvette. Fig. 76 on Plate IV. gives the midship section of the vessel shown in Fig. 52, whilst Fig. 77 represents a transverse section taken tovsrards the ends. The framing, both longitudinal and transverse, is partly continuous afid partly intercostal ; and will be described in such order as — it is thought — will enable the method of construction to be best understood. The flat keel consists of a single thickness of plate, 3| feet wide and y thick, extending continuously from stem to stern. Its connections, as also a description of the wood keel, will be given in Chapter VIII. The transverse frames are 3^ feet apart from centre to centre, those in wake of the protective deck being cut off thereby. Also, the framing above the deck differs from that below and beyond it. Every alternate frame of those below and beyond the protective deck extends continuously across the middle line of the ship, and is made up of a frame angle bar riveted to a reverse bar after the manner shown in Fig. 65 ; the frame angle bars are 8" x 3" x |" for the frames beneath the protective deck, and 6" x 3" x |" for those beyond it, whilst the reverse bars are 2^" x 2^" x ^" throughout. In Fig. 78 is indicated the lower part of one of these frames ; and the upper ends of those below the protective deck are formed into a knee for security thereto in a manner sim.ilar to that shown at a, Fig. 76, whereas towards the ends of the ship these frames extend continuously from topside to topside. When it is not possible to make the frame or reverse frame of a single length of angle bar the several pieces are connected together by covering angle straps. The vertical keel is partly continuous and partly intercostal, and extends from the stem nearly to the stern. The lower part is made of plates 11^' deep and |" thick, worked intercostally between the frames just described ; these are secured by a single row of rivets to a continuous plate 2 feet deep and |" thick, on the upper edge of which two angle bars, 3" x 3" x f " are riveted. Sections of the keel are shown in Plate IV., and a side elevation of a short length of it is given in Fig. 75, where i denotes the 94 Fic.75. O O O Q O O o O 0f;OO O OOOOOO oj- If ?[...Q - „.Q. O 00 jO o o O . o o c o ' o o c o o o o o o ;0 o o o ; o o o o ! o o o o ; o o o o 1 o 9 " b"o I d b'x o .o o .o n O O O O O O' _° intercostal plate fitted between the two continuous transverse frames /,/, and c represents the continuous portion of the keel. The butts of this continuous plate are j oined by double butt straps J" thick, riveted as indicated at h ; and the single angle bar, 3" x 3" x | , which secures the vertical and fiat keel plates together, has its horizontal flange continuoiis, whilst the vertical one is cut away at the transverse frames,/,/, to allow these latter to cross the keel. Connection is made between the intercostal plates and continuous frames by short pieces of angle bar 2^" x 2^" x t^", marked g in the figure. The first and second longitudinals from keel extend over about one- half the length of the ship, and are constructed somewhat similarly to the vertical keel, the main differences being that, in the longitu- dinals, lightening holes are cut, and double chain riveting is adopted for the butt straps of their continuous plates. Returning now to a description of the transverse frames below and beyond the protective deck, those placed alternately between the ones already described are shown in Figs. 76 and 77. They are worked intercostally from the vertical keel to the 2nd longitudinal, and continuously above it. From the turn of the bilge upward they each consist of frame and reverse frame angle bars, the former being 8" x 3" x |" for those below the protective deck and 6" x 3" x |" elsewhere, and the latter 2V' X 2V' X 1^". Below the bilges these frames are made of floor plates tSj" thick, projecting sufficiently above the longitudinals to take hold of two continuous angle bars 2^" x 2^" x t\, one of these being a prolongation of the reverse angle bar on the upper part of the frame, as indicated in Plate IV. The floor plates are connected to the vertical keel and longitudinals by angle bars 21" X 2V' X -i^" {see at d, Fig. 75), and to the bottom plating by 3" X 3" x I" angles ; and the one above the 2nd longitudinal laps I'taJtA IV Fi0 76 JudrftC^Liih 7 5 1 75 Fai-.-niqdon R" t Oociors Commons '^J!i4-.3 -91. To'farjt pobg*^ 9^^ 95 upon and is securely riveted to the frame angle bar, the junction being indicated at I in Figs. 76 and 77. The heads of these frames below the protective deck are also turned and secured to that deck as already mentioned. The transverse frames above the protective deck are, in the later vessels of this class, made of Z-bars, 6" x 3^" x 3" x |", and secured to that deck by a bent plate, as shown in the midship section. The earlier vessels had these frames each made of a frame angle bar, 8" x 3" x f ", and a reverse bar, 2^" x 2^" x -^", the heel being turned for security to the deck, similarly to the heads of the frames below. Lastly, the 3rd and 4th longitudinals from keel on each side of the ship are made up of plates worked intercostally between all the transverse frames. The former are incorporated with the coal bunker bulkheads for the greater part of their length, and are, like the 4th longitudinals, secured to the outer bottom plating by short pieces of angle bar. As can be seen from the figures, the inner edges of the 4th longitudinals are fastened to the inner flanges of the transverse frames by two continuous angle bars. The only part in the arrangement of beams and pillars calling for special remark is that in connection with giving support to the protective deck. Instead of working beams, longitudinal girders 3| feet deep and |" thick are placed above the deck and extend throughout its length. They are worked between the transverse bulkheads dividing the engine and boiler rooms, their ends being secured thereto by short angle bars. Each girder is attached to the protective deck by a single angle bar, whilst two bars are riveted to its upper edge, to which the wood lower deck is after- wards secured (see Fig. 76). This omission of beams below the protective deck gives additional room for the engines and boilers. The poop and forecastle beams are of angle bulb 5" deep ; those to the upper deck are tee-bulb 9" deep, and the beams under the lower deck — except where the girders mentioned above are utilised — are tee-bulb 7" deep. The pillars vary in size from 5" diameter and |" thick to 3" diameter and -f^" thick, the smaller ones being used beneath the poop and forecastle. 3. Of a Composite Sloop. On Plate V. are given two sections of the vessel depicted in Fig. 53 ; one taken amidships, and the other towards the ends. 96 The flat keel plate is 21 inches wide amidships and |" thick, and is continuous from stem to stern. The transverse frames are 20 inches apart throughout the ship, and are continuous from topside to topside across the keel, except those in "wake of the watertight deck over engines and boilers, which are cut off by it. The frames below the watertight deck consist of Z-bars 6" X 3^" X 3" X I", each having its lower end split into two angle bars and opened out to receive a floor plate -j^" thick which runs continuously from the turn of the bilge on one side to the corre- sponding position on the other. The upper and lower angle bars of the frames on opposite sides of the middle line are joined by covering straps, the former being butted at the middle line as at a Fig. 80, and the latter on one side of it as shown at 6, consecutive frames having this joint on opposite sides of the middle line. The lower angle bars are securely riveted to the flat keel plate, and the upper ends of these frames are connected to the watertight deck by bent bracket plates as indicated in the midship section. Above and beyond the watertight deck the frames are of Z-bars 4" X 3" X 2^" X f", the former being secured to that deck, at their lower ends, by bent bracket plates. As shown in Pig. 81, each of the frames beyond the watertight deck has its inner flange cut off below the bilges, and is converted into an angle bar 3" x 3" x f ", whilst on the upper edge of the floor plate, and extending from bilge to bilge, a reverse bar 2^" x 2^" x ^ is riveted. In the earlier composite vessels, frame and reverse frame angle bars were used where Z-bar frames are now employed. FicJ9. Fi<5.80 Tr f'lvc-e p 'xg e P6 PlaZe V. II Lower DecJe^ I I I I I I I I I [ Ju<)(itC°Lith.75i75,Firrmj(lonR«& DoctortCommonB. '^13*13 91) 97 The longitudinal framing resting upon the flat keel is known as the intercostal keelson ; it extends from stem to stern, and is made up of plates 18 inches deep and f "thick fitted between the transverse frames, and projecting sufficiently far above them to be connected to two continuous angle bars 3" x 3" x -I" "wtich are secured to the inner angle bars of the frames. The flat keel and transverse frames are joined to the intercostal keelson by half staple angle bars 3" x 3" x f" placed on each side. This arrangement may be seen in section on Plate V., and is given in side elevation in Fig. 79, where the intercostal plates are marked i, the half staple angle bars on one side b, and those on the other side c. The short angle straps to the upper frame bars (page 96) are lettered d. A longitudinal is worked on each side of the ship under the engines and boilers, and as far forward and aft as possible ; they are each formed of intercostal plates ■^" thick secured to a continuous piece of plate — which is riveted to the inner bars of the frames — ^by short angle bars, and to the outer frame bars by an angle which has its outer flange continuous, and its inner one cut away at the transverse frames. The steel watertight deck is supported on each side by a longi- tudinal girder g (Fig. 80), which is stiffened at intervals by brackets as shown at h ; also by the coal bunker bulkhead c, and, between that bulkhead and the ship's side, by several transverse brackets composed of plates and angle bars, one of these-being shown at d. The beams under the poop and forecastle, as well as those to the lower deck, are of angle bulb bars 5" deep ; the upper deck angle bulb beams are 6^" deep. The pillars in these vessels are small, varying from 2 to 3 inches in diameter, and are ^" thick. S 0. 11376. rcs JjTOV ^ Itoorru. WcObir Bailout BaJla-Sty Sbrres anjt Sftjdh I v, Hrdraulic Muyaiu ff'/.y oiine. %J2'^c^ Jl Fit^ines IT Etores , CapsUiTL aruJb noorr v 'WiHer Mdgaj^irijZ' Waierhghh Bulhisl Comparim JlydrauJic Fujtiping Room. ? SrrijiU. arrrv immmviwTi b _,\- Walcrtipht'^- ■ cirmpai'tmje/ll Wilier Waler BaJlnSl^^^^" D? Section Aft Section Amidships Section Forward Tofax-^cPuxte 111, JuddlC?Llth 75i75,Farnn9don R^ 8. Doctors' Commons. 2/3 4(3-5l) 115 of the water line ("between wind and water ") at the unarmoured ends are well sub-divided, as is also the part above the protective deck amidships, for the reasons given on pages 16, 44, and 47. The largest compartments are those containing the engines and boilers, and if one of the boiler rooms is damaged by a ram or torpedo, so that the wing passage and coal bunkei- bulkheads in its vicinity are also injured, the total amount of water which would find its way into the ship, supposing the compartments empty, would be about 700 tons. Since, however, a considerable volume is occupied by coals and machinery, the quantity of water entering the vessel would be considerably less than given above, and the vessel will heel over a few degrees only. This heel can be lessened, if thought advisable, by volantarily admitting water to the wing spaces and double bottom on the side opposite the damage. The capacity of one of the largest double-bottom compartments is about 40 tons. The extent of the sub-division in several other armourclads is exhibited in the following table, taken from the Manual of Naval Architecture previously mentioned. Ironclad Ships of Royal Navy. Watertiglit Compai'tments. Classes. Names. In Hold Spaoe. H-l 3 Total. Largest early types. Smaller early types. Largest recent masted < types. Smaller masted J types. ( Belted ships - 1 Mastless or J lightly rigged, j Rams - 1 Monitors - - ] Warrior Achilles Minotaur Hector Resistance Monarch Hercules Siiltan Alexandra Tem4raire - Invincible - Triumph Shannon Jfiilson - Devastation - Dreadnought Inflexible - Hotspur Rupert Gorgon Glatton 35 40 40 41 47 33 21 ■n 41 44 23 26 44 S3 68 61 89 2G 40 19 37 57 66 49 52 45 40 40 40 74 40 40 40 32 16 30 40 46 32 40 20 92 106 89 93 92 73 61 67 115 84 63 66 70 99 104 101 .135 58 80 39 97 S 11276 H2 116 TJnarmoured vessels are also well sub-divided, the Iris having- 61 compartments, the "C" class 77, and those of the Buzzard type 30. A general idea of the arrangements in the last two classes can be obtained from Figs. 52 and 53. The valne of efficient watertight sub-division has been abun- dantly demonstrated in actual experience, and one or two examples will suffice. In June 1880 the steam.ship Anchoria of the "Anchor"' line was struck nearly amidships by the Queen and had 28 feet of water in the damaged compartment, but the bulkheads kept her afloat. As an illustration of the usefulness of the inner bottom it may be mentioned that the Great Eastern, whilst proceeding to New York, touched upon a rock and tore a large hole in her outer bottom ; the inflow of water was, however, limited by the inner skin, and no inconvenience resulted, the vessel arriving at her destination in safety. The Iron Duke also damaged her outer skin by grounding-, but the injury was confined to two double bottom compartments. The accidental ramming of the Bellerophon by the Minotaur caused part of the wing passage of the former to fill, but the water was prevented from flowing into the hold by the wing passage bulkhead. Finally, the cases of the Canada and Hecla may be cited as two recent examples showing the utility of the collision bulkhead. The first collided with a barque near Halifax, and the second with a steamship in the English Channel, but in each case the damage done was confined to the foremost compartment. Attention will next be directed to the structural details of the bulkheads of the vessel shoAsoi in Plate X., those of smaller ships being constructed in a very similar manner. The plates are arranged in horizontal layers, and are from \" to -|" thick according to the size of the bulkhead and their position in it, the lower plates in anj' one bulkhead being slightly thicker than the upper ones. The largest transverse bulkheads are those forming the engine and boiler rooms, and their method of construction is shown by the part elevation and section given in Figs. 93 and 94, from which it will be seen that the plates are worked with flush edges and butts, the former being joined together by edge strips made of T bars, 4|" x 2|" x |", which also efficiently stiffen the bulkhead against buckling stresses ; these are shown at a. On the opposite side, angle bars h, 3" x 3" x fV", are worked vertically 4 feet apart, and these effectually co-operate with the edge strips in stiffening the bulkhead. This stiffness is often further augmented by fitting; T>hr/^ JfT rig . 93 Fig. 94. Section. THROUGH I. iw. Tig. 95. cb ^^3^^ tn To 7&ce pcufc //^. JuddiC'Lilh 75t?5 Fai-.-.ngdon R" & Doctors Common ^-'J** ■ •3- ■9/ 117 a strong lO-inch Z lt)ar at intervals, instead of one of the vertical angle bars. The butt straps fit between the T bars and, with these edge strips, are single riveted. The bounding edges of the bulkhead are united to the inner bottom, protective deck, &c., by an angle bar 4" x 3^" x iV marked c. The main transverse bulkheads worked towards the extremities of the ship have lapped edges and butts, and are stiffened by vertical angle bars only, spaced from 2 to 2\ feet apart. These bulkheads often receive good horizontal stiflEening by the abutment against them of the various decks and platforms ; for, unless a deck is plated, the bulkhead passes continuously through it, horizontal angle bars being provided on the latter to support the ends of the wood deck. The bounding edges of these bulkheads are secured to the outer bottom plating by single angle bars, and care is taken to make the bulkheads watertight where the continuous portion of the longitudinals passes through them. All the longitudinal bulkheads of this vessel are fitted between the transverse ones. That placed at the middle line of the ship has flush edges and butts, both edge strips and butt straps being single riveted. The stifEeners are all vertical, and are shown in the horizontal section through the bulkhead given in Fig. 95 ; they consist of small Z hars 3" x 2^" x 3" x y'i»" placed back to back as shown at d, and of intermediate angle bars e. The former are 4 feet apart, and act as pillars to the deck above. Additional stiffness is given to this bulkhead also by omitting a few of the vertical angle bars and substituting 10" Z bars. A single angle bar connects the edges of the bulkhead to the transverse ones between wh^'^h. it is worked, and to the inner bottom, whilst its upper eage is made watertight as indicated in Fig. 96. The bulkhead plates stop just beneath the deck beams, and are riveted to short pieces of plate / fitted between those beams ; these latter plates are joined to the angle bars g which closely fit against, and are riveted to, the deck plating and beams, and which can then be caulked. The wing passage and coal bunker bulkheads have their plates arranged on the raised and sunken strake system {see Plate III.), and their bounding edges are secured as just detailed for the middle-line bulkhead. They are stiffened by vertical angle bars 2 feet apart, the coal-bunker bulkheads having additional 10-inch Z bars at every 8 or 10 feet in lieu of the angle bars, to assist those bulkheads in resisting the pressure which would be caused by a 118 torpedo bursting in contact with the outer bottom. The wing- passage bulkheads in some recent ships have been made thicker (I") than usual for the same reason. The smaller bulkheads such as those to the shaft passages, shell and provision rooms, &c., are ^" thick, and have vertical angle-bar stiffeners 2 feet apart. Each bulkhead and watertight platform is tested for watertight- ness after it has been carefully caulked, either by filling the enclosed compartments with water, or by playing water upon the bulkheads from a hose. Any weeping of the rivets or caulking which results, is rectified, and the test again applied until perfect watertightness is secured. It is evident that all openings cut in watertight partitions for convenience of access to the various compartments must be capable of being made watertight. These openings include the manJioles cut in the inner bottom for access to the double-bottom spaces, the scuttles in the watertight platforms, and the doorways in the main bulkheads, as well as those to the various store rooms. Fig. 97 shows the method adopted for making a scuttle watertight. An angle bar to is riveted round the aperture, and to it the steel cover is fixed by the hinges h. On the under side of this cover, round its edges, a strip of india-rubber is worked, this being kept in place by thin strips of iron which run along its edges and are screwed to the cover. This is clearly indicated in the enlarged section through A B. To secure watertightness the india-rubber is forced tightly against the angle bar to by means of butterfly nuts, the screws of which are hinged to m and turn up into recesses cut in the ends of the hinges and in lugs I riveted to the cover ; the nuts are hove down until the india-rubber is well compressed. To lift the cover, the butterfly nuts are slacked back and the screws turned out of their recesses as shown a.i d.d ; the other nuts in the figure are indicated in their securing positions. Manholes are made watertight in a manner similar to that just described, except that the covers are sometimes jwrto/;fe instead of being hinged, and are oval in order to conform to the shape of the hole. Small screw plugs are fitted in manhole and watertight scuttle covers, the removal of which allows air to escape when the compartment below is being filled with water ; they can be easily replaced when the operation is completed. Large doorways are generally cut to afEord communication between the boiler-rooms, &c., as well as in the main bulkheads at other parts of the ships low down in the hold ; and since, in Platf^jm. Tig . 97 . A'Av.'www? :?;.^ 'i;. \' w^ !?^■^.^■;^M Enlarged Section HROUOH A . B . r. ^ India, rubber. Traru strip . EteVATlON Armour Deck Enlarged Section through a.b. Side Elevation OF Wedpe JuddiC? Lith 7Si75,Farrinj()on R^ i DsclorlCsmmafii "H* i : 9t ■p:rr- f^r," fIS 119 the event of a large breach being made in the side, the water would probably rise too rapidly in the compartment to admit of the doors being closed from below, if previously open, provision is made for shutting them from the first deck above water. These doors slide in either a horizontal or vertical direction, the fittings for one of the latter being given in Fig. 98. From the elevation, and enlarged sections through cd and ef, it will be seen that a frame F is worked all round the doorway, its sides, of the section shown, tapering from top to bottom so that from the thickness indicated in the section through c d, each side becomes reduced in thickness to that of the bottom part of the frame, (shown in the section through ef) where it joins that part. The part ^ g' of each side is at the same distance from the bulkhead throughout the length of the frame, and thus each forms, with a cheek C (bolted to the frame and bulkhead), a wedge-shaped groove in which the door works. This door consists of a wrought-iron plate, stiffened by "J" bars s, s, and secured at its edges to a wrought - iron frame W, the sides of which are tapered to the same extent as those of the frame F. Round the edges of the door, on the side opposite W, strips of brass are screwed as indicated in the sections, and thus, as the door descends in the grooves, these strips are forced against the frame F by the wedge-action of the door, watertightness being thereby secured. The door is actuated by a rod r, the upper end of which terminates at the first deck above water, whilst its lower end is jointed to a screw S, which works in the nut t attached to the upper part of the door. The rod is made square at its upper end, so that it may be turned by a spanner made to fit over its head. As the rod is prevented from moving vertically by the bearing k, the door must be raised or lowered as the screw revolves, the latter movement occurring when the spanner is turned with a right- handed motion. A deck socket is fitted to receive the upper end of the rod, and contains a tell-tale arrangement worked by the rod, which shows whether the door is open or shut. Fig. 98 also indicates the method of moving the same door from below. A spindle passes through the bulkhead and carries the pulley I, and the bevel-wheel h ; round the former, as well as round other wheels attached to the bulkhead, a chain passes as shown. The rod r is fitted with a cog wheel w, which, together with h, gears with a third wheel. Turning the pulley I by means of the chain thus causes S to revolve and work the door. A second pulley is often fitted to the spindle on the opposite side of the bulkhead so that the door may be actuated from either side. 120 In horizontal sliding watertight doors, the frame round the doorway is placed horizontally, and the vertical rod for moving the door has its lower end in a bearing fixed to this frame on its lower edge. The rod carries two cog wheels ot jnnions at different parts of its length, and each gears into a corresponding rack fixed to the door, so that the latter is moved as the rod turns. The bottom groove in the frame is liable to become choked by dirt, which would interfere with the closing of the door ; to. prevent this the groove is covered by a sill which is automatically removed by the door when being closed. When magazines, store rooms, &c., are not in use, their entrances are kept closed, hinged (or swinging) watertight doors being employed for this purpose ; similar doors are also used for those openings in main bulkheads which are situated high up, as there would be ample time to close them in the event of an accident. The method of constructing one of these doors can be seen from Fig. 99. A frame / surrounds the doorway and has a groove cut along its sides and ends, into which strips of india-rubber are received ; these are secured at their edges by narrow thin pieces of iron as indicated in the section through A B. The door consists of a thin wrought-iron plate with an angle bar d riveted round its edge, and is hinged at c to the frame of the doorway. To fasten the door in place, several handles marked h are emploj^ed. These turn in sockets fitted in /, and are so shaped as to press against vertical brass wedges on the door, when the latter is in place. Thus, the harder the handles are forced upon these wedges, the more will the india-rubber be compressed by the angle bar upon the door, and it is this compression which ensures watertightness. It will be noticed that the door may be secured from either side of the bulkhead, and also by turning the handles in either direction, since the wedges are tapered at each end, as will be seen from the side elevation of one of these given in Fig. 99. To allow of that edge of the door to which the hinges are attached being well forced against the india-rubber, the bolt hole in each hinge is made oval, so that the door is not rigidly bound at those hinges. It will be obvious that in the time of trial the value of the bulk- heads will depend upon the rapidity with which the doors in them can be efficiently closed, and hence the necessity for frequent drill and inspection to see that they are in thorough working order. In the Royal Navy this inspection takes place weekly. 121 CHAPTER XL PUMPING, FLOODING, AND DRAINAGB AKRANGEMENTS. The above arrangements provide for : — (1). Voluntarily admitting water into the ship, either to alter her trim, or heel, to increase the stability, or to fill the magazines, spirit rooms, &c., in the event of their being threatened by fire. These operations come under the head of flooding. (2). Clearing water out of the ship which has been allowed to enter voluntarily ; or which has accumulated through ordinary small leakages, by accident, or from damage done in action. Also for distributing salt water throughout the ship for wash deck, sanitary, and fire purposes, as well as for conveying fresh water from the storage tanks to the filters, galleys, &c. For these purposes pumps driven by steam, as well as smaller ones worked by hand, are used. (3). Collecting water which is to be pumped overboard into convenient places for this purpose. This is done by a number of pipes, which form the drainage system of the vessel. It will be noticed that the pumps are not relied upon to eject water from a ship as fast as it enters through a large leak, the inadequacy for such a task of the pumping power usually supplied to a vessel being easily demonstrated. If a hole is situated h feet below water, the velocity with which water enters the ship through it is given by the following rule : — Velocity = V h feet per second. Hence, if A represents the area of the hole in square feet, the amount of water passing through it in one second is 8 A s/ h cubic feet. The maximum quantity of water which can be thrown overboard in an hour by the armourclad most liberally supplied with pumps, does not exceed 4,500 tons, or 1^ tons per second ; and the above rule will enable an estimate to be made as to how large a leak, situated, say, 9 feet below water, can be just kept under by these pumps. If A represents the required area of hole, the quantity of water entering the ship in each second is 8 A -^ 9 = 24 A cubic feet. The amount of water got rid of by the pumps is 1^ x 35 cubic feet. Hence 24 A = IJ X 35. or A = 1-82 square feet. 122 Remembering that a successful ram or torpedo attack would probably result in making a hole some 20 or 30 square feet in area, it is seen how hopeless would be the case of a vessel when so attacked, if dependent upon the pumps alone for safety. This calculation shows how necessary is the -watertight sub- division described in the preceding chapter, care being taken in time of accident to isolate the damaged compartment from the others. As that compartment fills, the rate of inflow of the water lessens, and then a leak stopper of some description may be suc- cessfully got into position over the hole. If this can be accom- plished, the pumps can be set to work to empty the compartment so that any possible temporary repairs may be efEected from inside. Steam Pumps. — In arranging the steam pumping power of a ship the circulating jJumps, designed for use in connection with the condensers of the main engines, are fitted so that they may be employed in case of necessity for pumping water out of the ship, this water being then utilized for condensing purposes instead of drawing from the sea. For this reason the pumps are fitted with bilge suctions (pipes with their ends opening into the bilges, where the water to be pumped overboard is accumulated) through which water from the ship can be drawn into the condensers before it is discharged overboard. These pumps are now driven by independent sets of engines, this being preferable to the earlier plan of working them from the main engines, because, in the latter method, any reduction in the number of revolutions of the pro- pelling engines consequent upon an accident, correspondingly diminishes the jjower of the pumps. Again, when independent engines are used they can be placed high up in the engine room and considerably above the pumps worked by them ; the latter can therefore be kept going even when the engine room is partially flooded. The air pumps in some vessels have also been utilized for the same purpose. The other steam pumps ordinarily fitted in large ships are, the main engine bilge jiumps worked by the propelling engines and used for pumping moderate quantities of water from the bilges ; the main and auxiliary fire and bilge pumps, actuated by separate engines, and used for fire and wash declr purposes as well as for pumping from the bilges ; small " hand " pumps fitted so as to work from the main engines or by hand ; and, in some modern vessels, Friedmami's ejectors. The last are supplied in order to discharge rapidly overboard a large quantity of water, especially from the coal spaces at the extremities of central citadel ships, their com- parative immunity from choking rendering them very suitable for 123 that purpose. Each ejector consists of several nozzles with their axes on the same straight line, a small space separating each nozzle from the adjacent ones. To one end of the ejector a steam pipe from a boiler is led, whilst a larger discharge pipe is connected to the other. The ejector is surrounded by a strainer and placed in an fjector tank, in which is collected the water to be discharged OYerboard. On steam from the boiler being forced through the nozzles a partial vacuum is created, this being filled by water passing between the nozzles, which water is then forced overboard through the discharge pipe by the pressure of the steam. An ejector having a 4|" steam pipe and an 8" discharge pipe is said to be capable of throwing overboard 300 tons of water per hour. They are, however, very wasteful of steam and have not been fitted in the most recent ships. Hand Pumps. — For a long time the largest pumps worked by hand alone were those known as Downton pumps. They usually had three buckets working in the pump barrel, with a valve in each, and their principle of action was exactly the same as that of the common lift pump ; but, by having three buckets the flow of water was made continuous and they were m.uch more power- ful than one of the common pumps with the same lift. Those now used, however, are much simpler in construction, and doors on the outside of the pump enable the valves to be easily got at and cleared should they become choked from any cause. A skeleton sketch of one of the latest pumps used is given in Pig. 100. A solid watertight piston P is rigidly attached by a rod to a slide S, which is made to work up and down in the head of the pump by means of a crank (shown in side elevation) which passes through the pump, and can be turned by manual labour ; this operation causes P to move up and down in the pump barrel. In that barrel an orifice e is cut at its lower extremity, whilst a similar one / is formed at its upper part. The water to be pumped over- board enters through A, and its passage to the pump barrel is con- trolled by the valves a, b, c, d. Consider the action of this pump during one complete stroke, the piston being initially in its lowest position, as shown in the Fig., and the pump barrel being assumed full of water. As P rises, the water above it is lifted and forces its way through the valve d, and thence escapes through the delivery pipe D. But as the piston rises, a vacuum is formed behind it, which causes water to pass into the pump barrel through A, the valve a and the orifice e. When therefore P has reached the top of its stroke, the pump barrel is again full of water which, when the piston descends, is forced out of the valve h and thence- 124 to the delivery D. But, in descending, the piston creates a vacuum above it, in consequence of which water flows through A, the passage g, the valve c and the orifice /, so that at the end of the stroke the pump is exactly in the same condition as at the beginning, and the pump barrel has been completely emptied twice. It will be noticed that the piston of this pump is effective in both the up and down stroke, acting alternately as a lift and force pump ; it is thus superior to the Downton, which — ^having valves in its buckets — is effective during the up strokes only. The largest sized Downton pump ordinarily used was 9" in diameter, and was estimated to discharge 90 gallons of water per minute or 25 tons per hour ; this is also the power of one of the 9" pumps of the newer pattern illustrated in Fig. 100. In addition to the foregoing, smaller lift and force pumps are also supplied for use in connection with the galleys, wash places, &c. A large pump is made capable of drawing from several com- partments in one of the two ways illustrated by Pigs. 101 and 102. In the former, a suction-box or valve chest V is fitted beneath the pump and connected to the bottom thereof by the tail pipe shown. Pipes marked a lead from the several compartments into the valve chest, and their upper ends are fitted with screw-down valves, which can be opened and closed by the handles h ; the details of these valves will be given a little later. All that is therefore necessary to put any particular compartment in communication with the pump is to lift up the valve which is over the end of the pipe from that compartment, the other valves in the chest remain- ing closed. The second method of attaining the same result, shown in Fig. 102, consists in employing a deck- or suction-jJlate D, to the under side of which, at its centre, the tail pipe from the pump is attached. The upper ends of the pipes from, the several compart- ments are each led underneath the plate to a separate hole in the plate D, near its edge, and are usually kept covered by screwed caps marked b. On the upper side of the plate a goose-neck or arched pipe g is fastened at the centre hole, around which it can revolve ; and its length is sufficient to enable it to be screwed at its other end to any of the suctions, the covers h being removed for that p^^rpose. To enable the pump to draw through any particular pipe, the cover from that pipe is taken off and the goose-neck attached in its place, as indicated in the figure. When a deck- plate is adopted, it should always be well above the water-line, to prevent the possibility of water entering the ship through a Plat£^.rTTT Tig 100, Elevation of Pump iia— J J) v; \A L Tig. 103. Side Elevation OF Crank ZD ^ 2^sa Fi^. 101. "pi. ^li^ IPl fig. 104. xgr- Hg.l05. Jr» /fe<^ P"^^ yrrii:9don ^'. U P'clO' s Commoni Xiat.3/^1. To fijc^ P'^f- ^^^- 137 reverted to. Very high speeds have, however, been given to recent vessels, and as increase of speed greatly augments the pressure on the rudder, balanced rudders have in some cases been re-introduced to avoid the very strong and heavy appliances which would otherwise have been necessary for holding the rudder in position at extreme angles. Fig. 109 shows the general method of framing the rudder of an iron or steel ship. It is usually of wrought iron, and the head and front part of the frame are made in one piece, in which the pintles for carrying the rudder are formed. There are three of these pintles in the rudder shown, of which only the middle one passes completely through the corresponding lug on the sternpost. Into the ends of each of the others a small hemispherical point of hardened steel is screwed, and these bear upon similar points in the stern post, thus taking the weight of the rudder without pro- ducing much resistance to its rotation. The framing of the back part is joined to the front portion, and consists of an outer rim bent to the shape of the rudder, with two supports between it and the front edge, as shown by ticked lines. The rudder framing tapers in thickness from the front edge, as in the plan, and wrought iron plates are tap riveted to it in order to form the sides of the rudder, the space between them being filled up with some light wood, generally fir, to prevent the entry of water. For balanced rudders similar methods of construction to the foregoing obtain, it being observed that the only pintle used in a balanced rudder is placed on its lower edge, opposite the head. Cast steel is now often adopted for the rudders of iron and steel ships, and then the frame is often made in one piece. In vessels sheathed with wood, and coppered, the above materials could not be employed for these parts, for reasons already stated in Chapter YII. ; in the largest sheathed ships, therefore, the rudders are made entirely of gunmetal (or, in the latest vessels, of phosphor bronze), whereas the rudder of a corvette or sloop is generally of wood, with a gunmetal or phosphor bronze casting at its upper part to take the rudder head, although rudders made entirely of phosphor bronze are now being used in these smaller vessels also. A stuffing box is fitted where the rudder head passes through the stern of the vessel in order to secure watertightness ; one of these is shown in detail in Fig. 110. From the same figure it will be seen that the weight of a balanced rudder is taken inboard by the steel casting 0, which is rigidly attached to the rudder head and which bears upon, and slides in contact with, the sternpost at 6 b. This casting, and therefore the rudder, can 138 be locked in various positions, if required, by the locking pins L ■which pass through it and into the stempost. Instead of the sliding contact just mentioned, the weight of the rudder is some- times taken upon several small rollers to diminish the friction as much as possible. The lower pintle of the balanced rudder is simply for steadying purposes, it being received into a spur pro- jecting from the lower part of the stempost. The size, shape, and general arrangement of the gear for actuating the rudder differ greatly according to circumstances. Fig. Ill shows these details (excluding the steam steering engine and hand gear) in plan for a small ship. The rudder head A is clasped by the tiller, and the latter is prevented from slipping round the former by -cutting vertical grooves out of each in corresponding positions and fitting keys K therein. Any movement of the tiller must thus afEect the rudder, the latter being turned to star- board when the former is moved to port, and vice versa. The forward end of the tiller is made rectangular in section and of constant size for about 5 feet of its length, and passes through a Rapsori's slide, i.e., a metal block provided with a vertical spindle which revolves in the carriage C. To move the tiller from side to side, C is joined to the steam and hand steering gear by steel wire ropes which lead round the system of pulleys shown. At their forward ends these ropes are joined by a chain which passes round a horizontal wheel, the circumference of which is formed to take hold of the links of chain. By turning this wheel therefore by the steam steering engine, or the hand wheels, th.8 tiller is moved to starboard or port. As the tiller approaches the side, the carriage C slides forward along it, and thus causes the force pulling the tiller over to act at a greater leverage from the rudder head than would otherwise have been the case. When a vessel is fitted with a lifting screw, the necessary aperture renders the above arrangement inapplicable. Fig. 112 shows, in plan, a method sometimes adopted in such a case. The cross-head D is keyed to the rudder-head, and a second cross-head E, connected to the tiller, is secured to a pivot or auxiliary rudder-head C, the two cross-heads being joined by the parallel rods R. These rods thus communicate any movement of the tiller to the rudder to precisely the same extent as would have been the case had it been possible to connect the tiller directly to the rudder-head. Again, when a ship is provided with a complete protective deck, which is well below water at the ends, the steering gear is placed below that deck (where the vessel is necessarily narrow) for protective purposes. When this is the case, the width of the ship FJah'. Xim Fi0. HI /iSiSr.rTziii TilLii' 'y\V H vol 4--— of Ship Ti> ihucPo^e I'.i.t MiiJun^ a 0" FJg.112 r-.-ts^ Aperlure thr PropelLe-r Laie. _ r7S Fig. 113 I>cck n : M>~uv, l UJi-r 1 Jud()l.C?Lilh /Jl/S.fiT'ngdoii R' !. Docio.-s Commons Zl34-(3-9l! 139 is tiflnally insufficient to admit of an ordinary tiller being put over to the extreme angle desired, and then the parallel bar arrange- ment shown in Fig. 109 is generally adopted, or the screw gear sketched in Fig. 110. In the former an auxiliary rudder-head A is litted, supported by the structijre S ; and the fore end of the tiller passes through a carriage and block similar to that in Fig. 111. In this instance, however, the carriage is provided with wheels, and runs on a carriage-way C, worked transversely across the ship. Each end of the carriage is joined to a chain passing round a sheave at each end of C, and thence round the sprocket wheel W, which is worked by the hand and steam steering gear. The details of the apparatus shown in Fig. 110 are as follow : — A cross-head C (which is made in one with the part 0, already mentioned) is well keyed to the rudder and attached to the connecting rods R, each rod being joined at A by a single bolt to the sleeve S, through which a fixed guide rod F passes. Each sleeve is free to slide along the corresponding guide rod, and is firmly bolted to a nut N, in which works a screw attached to the shaft W. It will be noticed that one half the length of the screw is right handed and the other half left handed, so that as "W is revolved by hand or steam the nuts simultaneously approach towards or recede from each other, thus turning C, and therefore the rudder. The advantage attending the use of this gear is that, since a screw is practically non-reversible — that is to say, no likely pressure on the nut in the direction of the axis of the screw will cause the screw to turn — the rudder will remain in the position in which it is placed by the screw, even though it receives severe blows from the sea. No extra assistance is therefore needed at the wheels in heavy weather when steering by hand, and all possibility of the wheels flying back and so injuring the men is avoided. Also, when using the steam steering engine the effect of a blow on the rudder would only be felt by the screw and would not be trans- mitted to the engine. j\.uxiliary means are always provided for moving the rudder of a vessel in the event of accident to the main apparatus. For example, if the steering engine and ordinary hand gear of a ship usually steered by a tiller broke down, a relieving tackle would be employed on each side. One block of each tackle is secured to a shackle on the end of the tiller, and the other block is fastened to an eye plate at the side of the ship ; the tiller can then be moved to and fro by hand. Sometimes an auxiliary tiller is fitted, as indicated in elevation at A in Fig. 113 ; this 140 can be shipped into the rudder head and worted if the main tiller breaks down from any cause. When screw steering gear is employed, a tiller is also provided for use in case of necessity. It •will be seen in the plan of Fig. 110 that the rods R are prolonged to d in order that, if anything happened to the screw gear, these ends may be attached to a tiller (working round an auxiliary rudder head, not shown in the figure) after the rods R have been detached from the sleeves, thus producing an ordinary parallel bar arrangement. This tiller is fixed just above the screw gear. Sometimes provision is made for joining the ends d together after the bolts at A have been removed, the rods being then used as an ordinary tiller. In recent ships telegraphic communication has been electrically arranged between the various working or conning positions, the steering positions, and the rudder head. An electrical transmitter is placed at each working position, by means of which the com- manding ofBcer is enabled to transmit to any particular wheel elsewhere (say to an underwater protected steering position) the angle of helm which he desires. This order is received on a dial or receiver at the steering position, and the wheel is put over to the side and degree required. In order that the officer and the man at the wheel may both be sure that the desired angle of rudder has been obtained, an automatic electrical reply is arranged at the rudder head which indicates on dials or indicators at the conning and steering positions the actual angle of helm given. The steers- man moves his wheel in the required direction until the actual angle recorded on his " indicator " from the rudder head corre- sponds with the angle of the order on his " receiver " from the ofBcer ; at the same time the officer's " indicator " shows him the same angle, and satisfies him that the order is carried out. Switches are arranged at the various instruments, so that only those required to be in use are working at the same time. 141 CHAPTER XIV. PROTECTION OP SHIPS AGAINST GUN ATTACK. In tliis chapter it will only be necessary to notice the character of the protection given to all war vessels — ^in a greater or less degree — in order that they may resist gun attack, the measures taken to ensure their safety "when assailed by a ram or torpedo having been already pointed out in connection with the watertight subdivision of ships. It is needful to shield against excessive damage from this cause the parts securing buoyancy and stability to the ship ; and at the same time to preserve intact such vital parts as the machinery, magazines, and steering gear, as well as the big guns of large vessels. Buoyancy must be protected by preventing shot and shell from damaging the skin of that part of the vessel which is situated below water, whether the missile first strikes above the water line or, when the vessel is rolling, below that line for the upright position ; and the arrangements adopted for this purpose in the neighbourhood of machinery, magazines and steering gear, equally protect these parts also. To maintain sufficient stability, it is important to preserve uninjured a large proportion of the water-plane area and reserve of buoyancy, since any diminution of the former would lessen the metacentric height, and, therefore, the power of the vessel to stand upright ; and any reduction of the latter would lead to decreased stability at large angles of inclination. Protection is attained by employing thick iron or steel plates on the side of the vessel or in the form of protective decks, by utilizing the coal supply, or by a cellular subdivision of the vessel, these compartments being sometimes packed with some light material to exclude water should the divisions become pierced, and at others being empty or utilized for the stowage of stores. These systems will now be referred to in detail. Ships provided with thick side armour are known as armoured vessels ; several of these are depicted in Plate XVIII., from which 142 can be gathered the amount of protection afforded in each ship to the several parts enumerated above. The first English armourclads were the small vessels Thunderbolt, Erebus, and Tarrar, built during the progress of the Crimean war ; the thickness of armour employed was about 41". The larger vessels, Warrior and Black Prince {see Fig. 114), followed, and had a partial belt of armour il/ thick, this being capable of resisting projectiles from the heaviest guns then afloat. These vessels were open to the objections that their ends and steering gear were unprotected, and hence the Minotaur class were made large enough (10,700 tons) to carry a complete belt of armour from stem to stem ; this is 5^" thick amidships, but thinner towards the ends. About this period more powerful guns were supplied to vessels, and hence a corresponding thickening of the armour to resist them became necessary. To accomplish this, without greatly adding to the weight of armour, the " belt and battery " system was adopted. This consisted in having a com- plete belt at the water-line, and a central battery amidships, into which the fewer but heavier guns could be concentrated. This shortening up of the space requiring to be armoured in order to protect the guns, made it possible to increase the thickness of the armour used. The Bellerophon, Fig. 115, exemplifies this distri- bution of armour and armament, the former being 6" thick. The Penelope and Hercules are of the same type, the latter being pro- vided with armour 9" thick amidships, and 6" thick at the ends. Succeeding ships had guns placed in an upper deck battery immediately over the lower one, a good gun fire right forward and aft being thereby secured ; this is typified in the Alexandra, Fig. 116, which has 12" armour on the water-line belt — tapering to 10" at the ends— and 8" and 6" plating on the batteries. A 1^" iron deck joins the top of the comparatively low belt armour on each side, to give protection against high angle or depressed gun fire. The Temeraire followed, this vessel having her upper deck armament placed on turn-tables in two fixed armoured redoubts or barbettes, one near each end of the ship ; her water-line pro- tection is similar to that of the Alexandra. In the cruiser Shannon, Fig. 117, the guns are unprotected by armour, except by a bulkhead 8" thick, worked across the gun- deck forward to screen them against a raking fire ; and the 9" armour at the water- line is stopped 60 feet short of the stem, its ends at this point being joined across the ship by an armoured bulkhead. Forward of this, a 3" iron under-water protective deck was fitted, which was the first example of a now very general method of pro- tection ; a deck, 1|" thick, also joined the top of the belt 143 armour. The later cruisers, Nelson and Northampton, are similar to the preceding, except that an under-water deck (3" thick) and armoured screen bulkhead were given to each end of the ship ; and the deck above the armour was 2" thick. The growing size and power of naval guns caused further concentration of the armament of large ships, and the central battery gave way to the plan of placing the guns in revolving towers or turrets, by which means the heaviest guns can be easily trained through large arcs, and the same guns used on either side of the ship. Fig. 118 represents the Dreadnought, which has a complete belt of armour from stem to stern, and a central citadel rising above this in which the turrets are placed ; the maximum thickness of armour is 14". All recent first-class battle ships have their principal armament concentrated into turrets or barbettes, as exemplified by Figs. 119, 120, and 51 respectively ; but to resist the powerful guns now carried by vessels, it is necessary to provide thicker armour than was given to the Dreadnought. To do this without undue increase in the size of ship, the ends at the water-line are left unprotected by vertical armour, and the weight so saved is utilised to thicken the remaining side plating ; strongly plated under-water decks at the ends preserve from damage the spaces below them. Thus, the Inflexible, with a short water-line belt of the same length as the central citadel, has a maximum thickness of armour of 24" at the water- line, and 17" on the turrets ; the under-water deck at ends as also the deck over the citadel, being of iron 3" thick. The parts immediately underneath turrets must be protected by armour in order to shield the turret-turning gear ; but this is not so neces- sary beneath barbettes (except as regards the ammunition trunk) since the barbette itself covers the machinery for actuating the turn-tables. In barbette ships, therefore, a longer though narrower belt is possible on the water-line without increasing the weight of armour as compared with a turret ship ; this will be seen from Fig. 51, particulars of that vessel having been given in Chapter V. This allows of the guns being placed farther apart, and, being in separate protected stations, they are not liable to be all disabled by the explosion of a single shell, as is possible when they are placed in the same contracted central citadel, as in the Inflexible. The guns can also be carried higher out of the water than is possible in a central citadel turret ship, on account of the extra weight of armour which would be necessary to protect the turret-turning gear. In the Trafalgar the maximum thickness of armour is 20", and the greater displacement of that vessel enables 144 a longer citadel and still longer belt, as compared -with the In- flexible, to be given to it without undue sacrifice of other qualities. The secondary armament of this vessel, placed above the citadel, is protected against a raking fire by a 5" armoured bulkhead at each end of the battery, and against fire on the broadside by 3" armour, in addition to the 1" skin plating. The SSans Pareil and Victoria have a distribution of armour differing some-what from the pre- ceding. They are provided with one turret only, and have a partial armour belt similar to, but longer than, that of the Admiral class ; from the 3" armoured deck over the belt, at its forward end, a barbette or redoubt rises, and encloses the base of the turret, thus protecting the turning and loading gear. Unarmoured vessels having a protective deck — as the Mersey, Fig. 121 — are known as protected ships ; in the later vessels this deck extends from stem to stern, but in earlier ships it covers the engines and boilers only, as in the " C " (Fig. 52), and other classes. The earlier ships also had their deck entirely below the level of the water, whereas later ones have theirs at or above that level at the middle line of the ship for a good proportion of its length. The decks of such vessels are necessarily below water at the sides, as shown in midship section of Mersey, in order to prevent projectiles striking between wind and water from entering into the compartments below. Fig. 121 gives the thicknesses of the Mersey's deck, and the maximum thickness of the large pro- tected cruisers Blake and Blenheim is 6". In all vessels the armour was made of one thickness of plate until the time of the Inflexible. In that vessel the side and turret armour was made up of two thicknesses, separated from each other by a layer of wood ; this constitutes the sandivich system of armour-plating, and was adopted for the side armour of the similar, though smaller, vessels, A/rta; and Agamemnon. The turret armour of the last-named vessels, as well as all the vertical armour of later ships, is in one thickness only. The material employed for the armour of all ships down to the Inflexible and the similar vessels mentioned above, was wrought iron ; but on the turrets of those vessels, comjjound or steel-faced plates were introduced, and this compound armour has, since that time, been exclusively adopted for all thick armour plating. It consists of a back or foundation plate of wrought iron, faced by a layer of hard steel (see Figs. 123 and 124), the t^o being fused j'i(ji^jrmi. Warrior (9ZOO tonjs) I1_..J1 Kg. 114. y^ ■Inflexible ( II8SO i^jn^- ] n .:,___r]. -;S- ^^^fWl^ '^hell fio^m J-'ig 119. -Bintj^-r s SFnffine^ I fii'il-^n Maaaxi-ne- Bellerophon { 7SM tone ) Fig ns. Trafalgar' Pig 120 Alexan dr a ( 9500 tons ) Ti^.im. Shannon {.'^yjo toKs) Fi^. 117 Dreadnought (lOS;^^' toriir) Fig. 118, Merseiy.' n I Fig 121 'Jl • 4 Lrwoii'.f Tw/u'tfl i fjffdjcft' Knilers '^tig oitoies Mersey Plan of Protective DtCK. o\a \Ci/aW I^rp- ^^bEErr±n£5sinb^^ Mersey. Midship Section (Enlarged/ CoO'l' 2.130^ . i . 91 To face'-poffg' I44-. 145 together. The thickness of the steel is usually one-third the total thickness of the plate ; and the reasons leading to the adoption of this armour were as follows : — A hard plate was required in order to break up the shot on impact, and thus attention was early directed to steel ; plates composed wholly of that material were, however, found to crack very much, although the shot was broken up. On the other hand, plates of ductile wrought iron did not crack, but allowed the projectile to get through intact, and these facts led to the trial of compound armour, whereby the requisite hard surface was obtained in combination with ductile wrought iron, the latter preventing excessive cracking. These compound plates are at least 20 per cent, stronger to resist the attack of shot than the same thickness of wrought iron. It should be mentioned, however, that in recent years great progress has been made in the manufacture of armour plates composed entirely of steel, and these have been adopted in some vessels of the French and other navies, good results having been obtained on trial. Protective decks, which were at first of iron, are now constructed of mild steel. Between the skin-plating and the armour, a layer of wood iacking is placed, composed of teak ; this is shown in Figs. 63 and 73. It is well secured to the skin plating by flat-headed bolts, which are either screwed into that plating or have nuts on their points ; and its functions are to distribute any blow given to the armour over a large area, so that local injury to the framing may be lessened ; also to act as a cushion, thus deadening vibrations caused by a blow, which vibrations would injure the armour fastenings ; and, in con- junction with the skin plating, to prevent splinters of plate, shell, &c., from entering the ship. Its thickness is 15" in the Admiral class, but only 6" behind the 18" armour of the Victoria, and 4" in rear of the 20" armour of the Trafalgar. These two last-named vessels have the strong framing behind armour sketched in Fig. 63, and the wood backing is made as thin as possible in order to dim- inish the elasticity of the armour support, thus helping to procure that rigidity which has been found necessary for the supporting structure of compound plates in order to bring out their maximum powers of endurance ; the weight thus saved from the backing is utilised for the strong plate frames. Wrought iron plates were fastened to the ship by wrought iron conical-headed bolts, which passed through the armour, and had S 11276 K 146 nuts on their points hove up against the skin plating behind armour on elastic cup washers, as shown in Fig. 122. Each of Kg 122 Enlarged Section Through Washers. Plan of Cup Wa--her these washers consists of a hexagonal cup washer, marked a and h in the enlarged section and plan respectively, in which fits loosely a similarly-shaped indiarubber washer, covered by a thin plate washer ; it is against this latter that the nut of the bolt presses The elastic washer was introduced in consequence of the liability of the nuts to fly off under the jarring effect of a blow upon the armour, and has greatly lessened that liability. For a somewhat similar reason the shank is reduced at c to a diameter slightly less than that at the base of the screw thread. Before this was done, bolts almost invariably broke off at tbe thread when subjected to severe jarring stresses, since they were weakest at that part ; but by introducing a long slightly weaker part c, all the stress is taken by it, and the bolts only stretch. A bolt for compound armour is sketched in Fig. 123. Tt does not pass right through the plate, because any holes in the steel surface would render the plate very liable to crack if struck by shot. These bolts are of mild steel, and are screwed into the back plate from inside the ship by means of a spanner, which fits over a square projection ^S", formed on the 147 Fig 123. ■ aTV "^?^^ ^«