TO SHIP CONSTRUCTION. 225 



value of the completed weld, a welded ship is at present experimental in nature and 

 can therefore be more easily justified if designed along conservative lines. 



Exception may be taken to this statement in that we have no exact method of 

 determining the efficiency of a rivet in the hull of a riveted ship or the soundness and 

 homogeneity of a casting or welded forging, but the answer is that the accepted 

 methods of testing such work have been fitUy justified by experience and have proved 

 that the "guesswork" involved has been within the limits of the factor of safety al- 

 lowed for in the design which per se is the result of such experience. 



In preparing the design and arrangement of material for a welded ship, it is 

 considered desirable to keep the following practical features in view : — 



1. The work has to be properly closed up for welding. 



2. Overhead welding should be reduced to a minimum. 



3. As much welding as possible should be done on the ground or in the shops 

 where the material can be handled in the most favorable positions and where the dis- 

 comforts and inconvenience of protective equipments are least felt. 



4. In certain parts riveting may be more convenient and economical than 

 welding. 



Design for a Longitudinally Framed Electrically Welded Vessel. — Plates 25 

 and 26 show the midship section and shell expansion plan of a vessel designed by 

 Mr. J. W. Isherwood on the Isherwood System of Ship Construction. The dimen- 

 sions are 303 feet o inches B. P. by 42 feet 9 inches Mid. by 23 feet 4 inches Mid., 

 and the deadweight carrying capacity about 3,960 tons. The arrangement of ma- 

 terial and scantlings is much the same as in a riveted vessel of the same type, and 

 apparently the designer has given full consideration to the practical limitations in 

 ship construction, as well as to the limitations demanded by our present knowledge 

 of electric welding. 



The seams of side plating are arranged clinker fashion, and as many attach- 

 ments as possible are welded to the shell decks and inner bottom plating before erec- 

 tion in order to reduce the amount of overhead welding. In fact, from a close study 

 of the design, it will be seen that overhead welding is practically avoided altogether 

 in the case of the strength welds. 



The shell plates in this design are 84 inches in width and each plate has three 

 longitudinals welded to it before erection, the butts of the longitudinals being welded 

 at or near the butts of the shell plate. The numbers of shell plates and approximate 

 total number of linear feet of "full," "light,'' and "tack" welds are shown on the 

 shell plan. The system of construction lends itself to a large proportion of the total 

 welding being done prior to erection, and in working out the details no attempt 

 has been made towards an unreasonable elimination of bolting or riveting. 



For instance, the longitudinals are attached to the shell plating by service bolts 

 spaced 12 inches apart and are welded at the toe and heel by continuous light welds; 

 then the bolts are removed and rivets or welded studs fitted. In other parts, where 

 the minimum number of bolts necessary to properly tighten up the work approaches 



