AN ANALYSIS OF TESTS OF WATER-TIGHT BULKHEADS. 



85 



the apex of the bracket and will go to the zero point at the lowest rivet 

 on the overlap of the bracket. The maximum bending moment thus found 

 on the overlap will probably not be much greater than that at the middle 

 of the stiffener, but the shearing forces acting on the upper rivets of the 

 bracket must be very great. 



By the second form of bracket, Fig. i, c, the bar of the stiff ener is split 

 and intimately connected to the bracket, or some other form of equally 

 efficient connection is used, so that the bracket may be regarded as form- 

 ing an integral part of the stiffener. We need not in this case consider 

 the action of individual rivets; the stiffener carries the entire bending 

 moment BS at the end, as given by the curve PS, but being reinforced by 

 the bracket, the stresses will, with a well constructed bracket, not be as 

 great as at the apex of the bracket. Fig. 2 gives the curve of stress for 

 a stiffener bracketed as in Fig. i, c. 



FIGl. b. 

 I. CURVE or STRcsses. 



a. CURVE or MOMENTS. 



Hence by this form of bracket, being more certain in its action than 

 the bracket shown in Fig. i, 6, we are only concerned with the free part of 

 the stiffener between the brackets, and it must be of advantage to make, 

 these of such a height that the bending moments occurring at the apex of 

 the brackets, C and D in Fig. 3, shall be not greater than the maximum 

 which is found near the middle of the stiffener at O. The result of this 

 requirement is shown in Fig. 3, where the height of brackets, BD and AC, 

 required for its fulfilment is given in the different cases of bracketing. 



Curves of bending moment are shown both for uniform and for increas- 

 ing (triangular) load. For a partly uniform, partly increasing load the 

 curve would be intermediate between those given on the figure. 



