BY DAMAGE DUE TO COLLISION. 69 



the water has free access across the compartment. As the compartment fills, the 

 center of gravity of the vessel and contained water would fall as indicated by the 

 line marked "Locus of center of gravity" in Fig. 2, Plate 19. The center of buoy- 

 ancy would rise as indicated by the line marked "Locus of center of buoyancy." The 

 metacenter would fall as indicated by the line marked "Locus of metacenter." 



A condition of unstable equilibrium is indicated almost as soon as the water 

 "begins to enter the vessel. If the vessel remained upright so that the center of 

 ■gravity of the entering water would be on the center line of the vessel, this condi- 

 tion of instability would continue only until the compartment was one-third full and 

 a draught of 23.1 feet was attained. The metacentric height would rapidly in- 

 crease as additional water entered, and would become as much as i.i feet at 25.9 

 feet draught, when the water would cease to flow in, having attained its level. 

 The freeboard to main deck in this condition would be about 12 inches, and there 

 -would appear no reason why the vessel should not stay afloat in a perfectly calm 

 sea, although the margin of safety as regards freeboard would seem insufficient 

 except under ideal conditions of sea and weather. 



But the combination of unstable equilibrium and the inflow of water from one 

 side would surely cause the vessel to list toward the damaged side. This list would 

 continue to increase even after there was a positive metacentric height indicated 

 "by the diagram, owing to the fact that the center of gravity of the flooding water 

 would lie toward the low side, as obstructions would prevent a rapid flow across the 

 compartment. Fig. i shows that the list caused by the negative GM would cause the 

 main deck at the side to be submerged. When the main deck at the side is sub- 

 merged, the water will flow forward and aft along the deck and flood the adjacent 

 compartments, as well as further lower the metacenter. The list will continue to 

 increase, and the vessel will settle deeper in the water and finally sink. The time 

 elapsing from moment of damage to complete submergence would depend on a 

 number of circumstances, but recent collisions have shown that not over fifteen 

 minutes is sufficient. 



After complete submergence, there would be a righting moment due to the fact 

 that the materials of the houses, etc., would be of less density than the material of 

 the main hull. The cargo might be of such varying density as to give a righting 

 ■moment also. It is probable, therefore, that as the submerged vessel rests on the 

 bottom she would approach a vertical position with masts upright, or nearly so. 



From Fig. 3, Plate 20, it will be noted that the conditions are not far different 

 when No. 2 hold, 76 feet in length, is flooded. Permeability is taken at 64 per cent. 



It would appear, then, that in order to prevent this vessel from overturning 

 when one compartment is open to the sea by collision, it is necessary that the initial 

 metacentric height should be sufficient to prevent a condition of instability in any 

 stage of the flooding. For the particular vessel which has been investigated, the 

 initial metacentric height must exceed 1.4 feet by a margin sufficient to allow for 

 the upsetting moment caused by the fact that in the process of flooding an excess of 

 water will be on the damaged side. 



