176 THE LONGITUDINAL STRENGTH OF RIGID AIRSHIPS. 
as performance has reached a certain point, that it can be considerably extended or extended 
indefinitely. We know that this is not true, and I am hoping that some master mind of ex- 
perience and caution may be depended upon to guide in the tests of the great dirigible near- 
ing completion, that will safeguard it against early destruction or dying in infancy by increas- 
ing the severity of the tests to a point beyond anything that should ever be demanded. In 
this instance let us hope that ripe and mature judgment will take the place of enthusiasm 
and brainstorms in the matter of excessive demands or performance of our great product, 
so that she will live long enough to fulfill her mission in securing at least a part of the 
great mass of data that we need to make definite progress in an art that has such transcen- 
dent promise as does the one before us. It is certainly prima facie that these delicate struc- 
tires must receive treatment according to the “nature of the beast” to a degree probably not 
required in any other branch of the entire engineering art. 
ProFressor HovGAArD:—I am much indebted to the members who took part in this dis- 
cussion, especially Commander Land, speaking for himself, Commander Hunsaker and Mr. 
Burgess, all of the Bureau of Aeronautics. Commander Land states that the true solution 
lies between the method of shear and the method of bending. This is, of course, true in 
a certain sense and is pointed out in the paper in my conclusion No. 9, where I say that the 
shear method should be used locally in cases where the wiring does not harmonize with the 
distribution of the longitudinals; I ought perhaps to explain this more fully. 
By the bending method, we can determine the horizontal shear exerted by the wires in 
any panel. This should check with the horizontal component of the tension of the wires as 
determined by the vertical shear in the frame space, but where the wires are not properly pro- 
portioned to the longitudinals a discrepancy will occur. If the discrepancy is serious, it must 
be taken into account in applying the bending method. This happens in L-49 in the KM 
panels, adjacent to the heavy keel structure, where the wires are too light to secure a com- 
plete cooperation of the heavy keel with the rest of the ship in bending. A discrepancy, al- 
though less serious, is found also in the panels between the 4, C, and E girders. 
In illustration, consider the case of a beam lying in the hold on the bottom of a ship, 
which is exposed to sagging strains. If the beam is entirely free, it contributes, of course, 
nothing to the longitudinal strength of the ship, but if it be lashed down with transverse lash- 
ings to the bottom, so as to be forced to follow the vertical deflections of the ship, it will 
offer resistance as an individual girder quite independent of the ship girder. To the strength 
(moment of inertia) of the ship must, therefore, be added simply the strength (moment of 
inertia) of the beam itself. Now, imagine the beam to be further connected to the ship by 
horizontal or oblique lashings. It will then come to oppose the elongation of the bottom and 
will cooperate as an integral part of the ship in resisting the bending to an extent deter- 
mined by the strength of the horizontal lashings. 
In an airship like L-49, analogous conditions exist, inasmuch as the keel, consisting of 
no less than six longitudinals, forms a heavy triangular girder, which by the transverse frames 
is forced to follow the vertical deflections of the rest of the ship, but which is in shearing 
connection with it, only through the relatively light KM wires.’ If we calculate the horizontal 
shear forces in the KM panels required by the bending method, we shall find that it is much 
greater than that actually obtained by the vertical shearing deflection in any given frame 
space. It follows that the keel cannot cooperate in a complete manner with the rest 
of the ship. 
