The validity of the assumption made that the ordinary theory of bending is applicable to the case of a ship was tested and con firmed in 1903 by experiments made on H.M.S. "Wolf" by Sir John Biles. (See Trans. I.N.A. 1905.) The experiments were made both in still water and at sea, the extension or compression of the structure at various points being measured by strain indicators. As a result of these experiments it may be said in general that the stresses corresponding to any particular external conditions closely agree with those calculated on the usual theory of bending; but the waves encountered during the sea trials were such that the maximum stress then obtained was considerably less than that in the condition assumed in the standard calculations.
Several refinements for calculating stresses among waves have been suggested, none of which have come into general use. Some of these are the Smith correction (Trans. I.N.A. 1883), and the modifications due to heaving and pitching and rolling oscillations (see Trans. I.N.A. 189o, 1896, 1898). In addition to the direct stresses resulting from longitudinal bending, shearing stresses are experienced which in some cases are appreciable. Such stresses reach a maximum at or near a quarter of the length from either end, and at the height of the neutral axis. It is sometimes neces sary to give special consideration to the thickness of plating and the connections in this neighbourhood.
Stresses due to transverse bending may be of importance in certain cases ; and some means of comparing the transverse strength of vessels is desirable. In a ship well sub divided by transverse bulkheads it is difficult to determine how far the stresses at intermediate sections are influenced by the neighbouring bulkheads, positions at which transverse stiffness is very great indeed. But by taking a section midway between two bulkheads assumed far enough apart to have no influence on the section considered, a method of comparison has been worked out. For further information reference can be made to papers by Dr. Bruhn in Trans. I.N.A. 1901, 1904 and 1905.
In addition to the stresses due to longitudinal and transverse bending, there are numerous local stresses for which provision must be made. These include stresses caused by water pressure, heavy blows from the sea on side, deck and upper works, reactions of moving parts of machinery, blast from gun fire, etc. The supports provided are determined by experience and by the requirements of the particular vessel.
It is important that discontinuities and abrupt changes in the distribution of material should be avoided so as to maintain as uniform a distribution of stress over any section as possible. Where large openings have necessarily to be cut, "compensation" is introduced by adding material near the edge of the opening.
A ship like any other elastic structure has a natural period of vibration ; and when synchronizing vibrations from whatever cause are impressed on the ship very marked vibra tion may result. The principal causes of vibration in a steam ship are :—(1) The reciprocating parts of the engines, if un balanced cause vibrating forces and couples in a vertical plane. In twin screw ships torsional oscillations in transverse planes may also result when the engines are working in opposite phase. (2) The rotating parts of the engines cause vertical and horizontal oscillations ; and variation in crank effort also tends to produce torsional oscillation particularly in single or two-cylinder engines. (3) Unbalanced propellers produce vibration principally at the stern. (4) The uneven resistance which a propeller experiences during its revolution gives rise to vibrations. This condition re sults when (I) the propeller blades pass too close to the hull; (2) when the propeller breaks the surface of the water; and (3) when the flow of water to the propeller is imperfect, due either to "cavitation" or the screening effects of shaft and propeller supports. (5) Various items of auxiliary machinery set up inde pendent vibrations.
The adoption of the steam turbine has eliminated many of the causes of vibration referred to above. In general it can be said that vibration in a ship can only be avoided by removing its cause; it not infrequently happens that the addition of further stiffening results in increased vibration.
Reference can be made to papers in Trans. I.N.A. of Dr. 0. Schlick (1884 to 1901), of Mallock (1895) and of Nicholls (1924).
The information available on the steering and manoeuvring qualities of ships is largely due to the results of trials with British warships. These include observations of the paths when turning under different angles of helm, at various speeds, with and without assistance from the propellers, and with variation in certain features of the hull which influence steering.