The distribution of material here assumed almost exactly answers to that of maximum strength when the upper part of the transverse section undergoes a tensile strain, and that of the lower part a compressive strain, which, in fact, is the strain most likely to injure the ship when afloat. For, in this case, the distance of the upper flange from the neutral axis is 12 feet, while that of the lower flange is 11 feet; and if the ultimate compressive strain per square inch at the lower flange be 17 tons, the tensile strain per square inch upon the upper flange 17 x 12 will be or 20 tons nearly, which is about the ultimate

11

resistance of wrought iron to extension.

It may not be practicable in all cases to construct iron ships with such a large section of iron at the upper deck, yet such should be the distribution of the material in the section in order to secure a maximum strength with a given amount of material.

9. In the distribution of the material there is another consideration of some importance, and that is, that all bodies in the form of beams, whether hollow or solid, follow the same law as regards a transverse strain, viz., that in a beam uniformly loaded, the strains are always greatest in the middle, and progressively diminish to the points of support at either end. These facts are self-evident, and show, in the case of an iron ship, that the same thickness of plates is not required when working from the centre at midships to the stem and stern. In fact, they should taper or be reduced in thickness according to a certain ratio of their distances from the centre till they reach the extremes at each end.* Theoretically this is

* When a ship is supported at the extremities, and loaded uniformly throughout its length, the moment of the force tending to produce transverse fracture at any point varies as the rectangle of the parts into which this point divides the distance between the two points of support, and the strength of the section at any part should vary according to the same law. In a tubular beam, having its material accumulated at the top and bottom parts of the section, and with a uniform depth, the strength varies as the area of the section, so that, in this case, the area of the section should vary as the rectangle of the parts into which the point (where the section is taken) divides the distance between the supports. If such a beam be divided longitudinally into four equal parts, then the section of the material at the part midway between the centre and the extremity should be three-fourths of the section at the centre. And so on to other cases.

true, but in practice we have to consider how much the thickness can be reduced without danger to the structure, and in general we may here observe that the reduction should not exceed one-third between the centre and the two extremes. Or, in other words, if we assume the strakes or sheathing plates of the bottom and round the bilge to the height of the interior floor, or one-fifth of the depth, to be seven-eighths of an inch, it then follows that their thickness may be safely and progressively reduced to five-eighths thick towards the bow and stern. The same reduction to five-eighths thick may be made from that point, one-fifth of the depth, to the neutral axis of transverse strains, or about half-way up the ship's side, when they should again increase to seven-eighths thick for the top strakes at the deck, on each side where they have to perform the office of stringers and columns under the action of the two forces of tension and compression.

10. From these remarks, it is obvious that a careful distribution of the material is a condition of considerable importance in ship-building, and although it may be necessary in some constructions to deviate somewhat from the absolute rule, yet it is nevertheless essential that the law of strains should be carefully observed, and weak parts efficiently guarded against.

11. It will not be necessary to give drawings in illustration of these statements, as we apprehend the question to be sufficiently explicit to enable the iron ship builder to proportion the parts in the ratio of the strains, and to afford to the ship, as a whole, ample powers of resistance to the forces by which she may be assailed; care, however, being taken to provide for wear and tear, oxidation, and all those other influences which tend to weaken the ship.

71

CHAPTER VI.

THE COMBINATION OF IRON AND WOOD IN THE CONSTRUCTION OF SHIPS.

1. THE use of iron, as a material for ship building, suggested several modifications in the construction of ships built entirely of timber. Amongst others was the introduction of iron for the frames and ribs, and to these was screwed the wood sheathing, varying in thickness from four to six inches, according to the size and strength of the ship. This combination was for some time highly appreciated, but, when the properties of the two materials are considered, it will not be difficult to prove that the system is not an eligible one, and, for sea-going ships, is utterly at variance with sound principles of construction. It may be desirable in some cases to have vessels of this construction, such as lake or river boats, but for large ships intended to navigate the open sea, this construction cannot be recommended either on the score of economy or safety.

2. Numbers of these vessels-some of them of large tonnagehave, however, been built; and, in so far as regards fouling, they are superior to the iron-plated ships. As respects strength, they are nevertheless exceedingly defective when compared with those entirely composed of iron. Plate stringers have been introduced, and riveted along the ribs at different parts of the sides, decks, and bottom, but their position has little or no tendency to strengthen vessels of this construction. The following section, figs. 40 and 41, exhibits the principle on which these vessels are constructed, and, comparing this with that of a vessel of similar dimensions composed entirely of iron, we arrive at the conclusion that the method of wood sheathing will scarcely bear a comparison with that of closely riveted iron plates. To prove this, it will be necessary to refer to experiments-1st, upon the comparative strength of iron and wood to a tensile strain; 2nd,

upon their comparative resistances to compression; and, lastly, to their comparative powers of resistance to pressure.

TABLE I.-COMPARATIVE VALUES OF IRON AND TIMBER SUBMITTED TO A TENSILE STRAIN.

3. The foregoing table represents the ultimate resistance of the best qualities of iron and timber to a tensile strain, and from which it will be seen, that box and ash stand at the head of the list for strength, and that Spanish mahogany, oak, sycamore, and birch are the next in order, after which red deal or pine the timber chiefly employed in sheathing ships-is only 500 lbs. per square inch short of the last three. Selecting, however, English oak as the best for comparison, we have the two materials in the ratio of their relative strengths as 1: 3-84; or, in other words, iron is a little above 3 times stronger than oak. But in some other experiments, to which we shall hereafter refer, the timber is shown to be not more than one-fifth the strength of iron. These experiments, although bearing directly upon the resisting powers of iron and wood, do not apply to the two materials as ordinarily used in the construction of the sheathing united to the hull and iron frames of ships. To apply the comparison, it must be in the shape of two vessels, one of iron and another of wood, of similar dimensions.

4. The comparative strength of iron and wood, in regard to their respective powers of resistance to a tensile strain, is therefore in the ratio of the strength of the different sorts of timber, as given in the above table. Red pine is probably the best description of timber for sheathing, in combination with iron

frames. With 6-inch planking-provided there were no joints, and the ship was covered with a homogeneous mass of jointless timber-it would then be equal to plates of about 1 inch thick. But as this is not the case, and as the timber joints are not united, either longitudinally, vertically, or transversely, we must depend on the fastenings to the angle iron ribs, to which the sheathing is bolted, as the only power by which resistance can be offered to strain.

If we carefully examine this construction it will be found, the combined planking being separate from the ribs, that the fracture would be in the direction of the black lines a a a,

fig. 27, and the tendency would be to separate the end joints and twist the cross ribs, as shown at c cc, in fig. 29, and comparing the strengths with iron, as derived from experiment in the following table, it is evident that the strength of vessels of this description, as they are now built, is very inferior to those composed entirely of iron, where the whole of the sheathing, when securely riveted, is homogeneous, and presents greatly increased powers of resistance to every description of strain to which sea-going vessels are subjected.

On this question it will be noticed, that in the iron ship we have the whole of the plates united by rivets in every direction, and to which are secured ribs and frames again united by rivets, forming an unbroken covering of a solid sheet of iron. Now it is well known, from the results of former experiments, that the strength of iron plates riveted in this manner is reduced from 50,000 to 30,000 lbs. per square inch-as in the case of steam boilers arising from the quantity of iron punched out for the reception of the rivets. This being the case we must reduce the strength of the construction to that proportion. Bearing these facts in mind, we have now to ascertain the strength of the composite form, where the wood sheathing is not united in the joints, but on the contrary is driven asunder