July 25, 1907] 



NA TURE 



305 



from time to time b} Lloyd's Register Society. Classifi- 

 cation at Lloyd's is so important in the mercantile marine 

 for purposes of insurance that the design of a ship's struc- 

 ture is usually little, if any, more than compliance with 

 Lloyd's rules and tables. These rules have been modified 

 as ships have increased in size and varied in type ; and 

 when exceptional ships not directly provided for by the 

 rules have to be classed, the structural design is specially 

 dealt with by Lloyd's; but the governing principle through- 

 out is e.\perience of the behaviour of ships at sea. 



Lloyd's Register Society has also done, and is still 

 doing, much in the way of scientific research. It has 

 a highly trained technical staff which has conducted and 

 published some of the most valuable investigations yet 

 made of the structural strength of ships ; but the general 

 problem of how to obtain the requisite strength of struc- 

 ture with the minimum weight of material is extremely 

 difiicult of approach from the scientific side. The usual 

 calculations of structural strength are based upon still- 

 water conditions. The most important are those which 

 relate to longitudinal strength', because the greatest stresses 

 that can come upon a ship are in the longitudinal direction. 

 In these calculations the structural portion of the hull is 

 regarded as a steel girder supported over the whole of its 

 length by the upward pressure of the water. 



In order to approximate somewhat to the worst conditions 

 at sea, the ma.ximum stresses at the top and bottom are 

 calculated for two hypothetical cases of support upon a 

 wave surface. The surface usually taken is that of a 

 trochoidal wave of the same length as the ship, the height 

 of which is one-twentieth of its length. The vessel is first 

 considered in equilibrium upon a wave-crest with her bow 

 and stern in the adjacent hollows, and next to be sup- 

 ported at the ends upon two wave-crests with her midship 

 part in the hollow between them. The whole system of 

 wave- water is supposed, for the purposes of the calcula- 

 tions, to be fixed for the moment, and the ship to be 

 floating upon it in statical equilibrium. 



It is not known how nearly the results given by calcula- 

 tions, which rest upon assumptions that differ so widely 

 From the real circumstances, correspond with the maximimi 

 .stresses really brought to bear at sea, but it is certain 

 that they are often much in excess of the truth. In the 

 new big Cunarders, l^iisilaiiia and Mauretania, the limit- 

 ing stress accepted by Lloyd's as determined by calculation 

 was 10 tons per square inch for mild steel the ultimate 

 tensile strength of which is 28 to 32 tons per square inch. 

 This gives an apparent factor of safety of only 3. 



Many vessels have been running for years in which the 

 figures, obtained by similar calculations, for the maximum 

 stresses would amount to to tons per square inch. This 

 must be largely in excess of the truth, and it is impossible 

 to say exactly by how much. 



The quantitative values of the calculated stresses are thus 

 j'Xtrem.-ly doubtful. Even in comparing them with figures 

 obtained in a similar way for other ships, it is necessary 

 to be careful not to press the comparison too far. 

 Attempts have been made to measure the actual stresses 

 ;it se;i upon portions of a ship's structure bv means of 

 strain indicators. Extensive experiments were carried out 

 in H.M.S. Wolf a few years ago by an .Admiralty Com- 

 mittee with Stromeyer's indicators, which gave some 

 interesting results ; but very little real progress has yet 

 been made towards a quantitative solution of the strength 

 problem. 



The Great Eastern proved, by her Atlantic voyages to 

 New York and Quebec, and her subsequent experiences 

 in the trying work of cable-laying in the Atlantic, that she 

 ■was quite strong enough for anything required of such a 

 ship; and if we compare her structure with that of the 

 standard ship of her dimensions and tvpe to-dav, which 

 »=mbodies the results of fifty years' more experience than her 

 designer had at command, it appears very remarkable. Sir 

 W. \Vhite came to the conclusion, which I believe is right, 

 that after making full allowance for features of modern 

 designs, that involve additional weight, which the Great 

 Eastern did not possess, her structure was lighter than 

 that of the corresponding ship of to-day, although the ship 

 of to-day is built of steel ^o per cent, stronger than the 

 iron plates of the Great Eastern, and the riveting of the 

 edges and butts of plating is much more extensive and 



NO, Tq69, VOL. 7'S] 



efiicient, and is performed by hydraulic power in those parts 

 where strength is most important. 



The difference in principle between the two designs is 

 so great, and the comparison of the weights of material 

 they require is so much in favour of the Great Eastern, 

 that there certainly seems to be a case for careful investi- 

 gation, and for seriously considering whether a radical 

 change in the structural design of large ocean liners might 

 not be made with advantage. Novel structural arrange- 

 ments are constantly being introduced into the design of 

 cargo-steamers in order to give large open holds and to 

 facilitate stowage. Some of these are now being built of 

 large size and depth, with only a single strongly plated 

 deck at the top, and there seems no reason why this 

 principle should not be applied to large passenger vessels. 

 Any saving of weight thus effected would not only be a 

 saving of cost, but would better enable the difficulties of 

 draught of water in harbour and docks for the largest 

 ships to be overcome. 



The problem of speed has alwavs been a verv vexed and 

 difiicult one, and there is none which has caused more 

 trouble, or given rise to more fallacies in theory and errors 

 in solution. I cannot even direct attention to the 

 numerous theories and the various approximate formulas 

 that have been invented and employed from time to time 

 for explaining and solving the speed problem. These 

 formulas are generally so restricted in their range of 

 application, and require so much knowledge of their 

 limitations and the conditions under w-hich they can alone 

 be relied upon for results that will be approximately near 

 the truth, as to prove dangerous traps to the unwary and 

 ill-informed. The man who can use these intelligently and 

 safely, and with full knowledge of their limitations and 

 their tendencies to error, is able to deal with the speed 

 problem much more completely and effectively — and I shall 

 confine my remarks to the way of doing that. 



The practical solution of the speed problem was effected 

 by the late Mr. William Froude when he discovered the 

 law of similitude or comparison which enables the resist- 

 ance of a model, as ascertained by e.xperiment, to be used 

 for calculating the resistance of another model upon a 

 different scale, or that of a full-sized ship of similar form. 

 His analysis of the separate elements of resistance, show- 

 ing that the two great ones, friction and wave-making, 

 varied independently of each other, and the latter in a 

 very irregular manner, explained why simple approximate 

 formulas are so untrustworthy. 



What is wanted for the practical purposes of a designer 

 is the means of ascertaining the resistance of a ship of 

 given dimensions at any desired speed ; and also of readily 

 determining the precise form or degree of fineness of 

 under-water body that would en.able the maximum of carry- 

 ing power to be obtained at a moderate rate of fuel con- 

 sumption. It is one thing to know exactly what power 

 is required to give a ship of given dimensions and form 

 the speed asked for or promised, and quite another to 

 determine what are the dimensions, form, and degree of 

 fulness that will give the maximum passenger and freight- 

 carrying capacity with moderate engine power and ex- 

 penditure of fuel. 



In order to exhaust the problem of the best form of 

 ship to meet the requirements of any particular trade or 

 service, considerable investigation is required. This can 

 only be made satisfactorily by testing the resistances of 

 models in an experimental tank upon the late Mr. Froude's 

 system. 



That method is unfortunately impracticable, however, for 

 ship-designers in this country, because there is no experi- 

 mental tank here available for general use. The very few 

 that exist belong either to the .Admiralty or to private ship- 

 builders, and are confined exclusively to the work of their 

 respective owners. I have had experiments made occasion- 

 ally for my own purposes, but had to go abroad for 

 them. The experiments required by Mr. Yarrow for 

 his valuable investigations into the effect of shallow water 

 upon speed were made in the North German Lloyd tank 

 at Bremerhaven, where other experiments have also been 

 made for him. .\ British shipbuilder can only get such 

 experiments made by setting up an independent establish- 

 ment for himself or going abroad. Now an experimental 

 j tank, with its equipment and a competent staff for work- 



