THE LONGITUDINAL STRENGTH OF RIGID AIRSHIPS. 145 
of an ordinary ship if openings were cut in the shell plating between the frames and Jongitu- 
dinals as indicated in Fig. 1, where a thin diagonal strip of plating, incapable of resisting 
compression, is left in each panel. Here, again, the lines of compressive stress at the neutral 
axis will be directed along the frames and longitudinals. This peculiar distribution of the 
stresses will locally influence the strains in the structure and may, for instance, cause a longi- 
tudinal to be in compression between two joints, which themselves are subject to a tensional 
stress across a transverse plane through the center of the joint. This anomaly exists where 
strong shearing is combined with a small bending moment. It follows that only at the joints 
are the stresses distributed as in a continuous plated structure, and if we want to apply the 
ordinary theory of bending to determine the stresses, it appears from these simple consider- 
ations that it is best to consider sections at the frames taken through the center of the joints. 
II. METHODS OF CALCULATION. 
In order to calculate the stresses in the girders and in the wires, it is necessary first to 
consider, in a general way, the mode of loading. 
When an airship is floating in statical equilibrium, the straining actions are due solely to 
the buoyancy and the weights which are readily calculated. The buoyancy is very evenly 
distributed (about as in a submarine) and part of the load—viz., that due to the hull—is 
distributed practically in the same way; also the consumable weights: fuel, ballast, ammuni- 
tion, etc., are fairly uniformly distributed, but the remaining load resides chiefly in the cars 
with their machinery, the weight of which is applied in a rather concentrated manner sus- 
pended under the hull. The consequence is that in the vicinity of the cars the shearing has 
maximum values at points where considerable bending exists at the same time. In one or 
more places, on the other hand, the conditions of loading may come to resemble those existing 
in the axle of a carriage, where pure bending, practically without shearing, exists between the 
bearings. The load and buoyancy are so arranged that the ship is normally in hogging con- 
dition so as to ensure that ordinarily only the lower longitudinal members shall be in com- 
pression. This is desirable because failure of the longitudinals is much more likely to occur 
in compression than in tension, and the heavy keel structure which runs along the bottom of 
the ship is better able to stand compression than the top girders. 
