AN ANALYSIS OF TESTS OF WATER-TIGHT BULKHEADS. 95 



It will be seen from this table, that for the double stififeners of the 

 type here tested it would be safe to allow a stress of 1 2 tons per square inch 

 if same assumptions are made as in the present calculations. 



For single stiff eners a permissible stress in pure bending of 10 tons 

 per square inch has been chosen, and this figure has been used as basis 

 for the accompanying tables. As seen from the table of stresses, tests 

 Nos. II and 12, where the stiff eners were fixed at one end only, much higher 

 stresses may be reached without any breakdown taking place, but it will 

 be noticed that the permanent set is considerable in these cases, and it 

 appears preferable to bracket both ends and to allow only the more moderate 

 stress. It is argued that since stiffeners so constructed have proved them- 

 selves to be not only safe but to have a very small permanent set, we shall 

 arrive at a satisfactory construction of stiffeners in the design of new bulk- 

 heads, even if of very different type, if we make similar assumptions, and 

 if we keep within the stresses found in the present analysis. 



III. General Rules for the Construction of Water-tight Bulkheads. 



While the following rules are essentially a resume of the foregoing, 

 they are based also largely on tests, experience and practise, which space 

 does not permit to reproduce or discuss in this paper. Among the material 

 drawn upon may be mentioned a great number of tests, too incomplete 

 or of too complex a nature to permit an analysis, and further the current 

 practise of various foreign navies and the rules of the Classification Societies. 



The rules here given apply only to ordinary water-tight bulkheads, 

 not to deep tank or to oil-tight bulkheads. 



1 . A pure system of uniformly spaced vertical stiffeners and horizontal 

 strakes of plating should be used, except where the horizontal extension 

 of the bulkhead is smaller than its height, and where at the same time 

 the brackets of horizontal stiffeners can obtain a rigid attachment, in which 

 case a pure system of horizontal stiffeners should be used. This last con- 

 dition is fulfilled only in few cases, such as in collision bulkheads. In the 

 majority of bulkheads vertical stiffeners are the most efficient as far as 

 strength under water pressures is concerned. 



2. The thickness or weight per square foot of the plating may be deter- 

 mined on the basis of its depth below or its height above the water-line, 

 which corresponds to the deepest draught to which the ship may reasonably 

 be assumed to be immersed in damaged condition. As a general rule it 

 is proposed to make the weight of the strake immediately below this water- 

 line not less than 7 pounds per square foot, and to add one pound for each 



