philosophy not adhered to at the Model Basin, the starting point still goes 

 back to rational methods of analysis for the elastic-instability problem. 



At the outset, it is important to state categorically that, contrary to 

 the belief of others, elastic instability of stiffened cylindrical shells under 

 hydrostatic pressure is not obscured by the "snap-through" mechanism. 

 This is in contrast to the problem of cylinders under axial compression or 

 torsion loading where, as has been pointed out by Thielemann in an 

 excellent paper on the nonlinear theories of buckling of thin cylindrical 

 shells, even for the case of no imperfections, the effects of large defor- 

 mations are important. 



A detailed investigation of the "Durchschlag" problem by Kempner 

 et al for cylindrical shells of short length which are perfectly circular 

 and initially stress free has provided theoretical confirmation of the ade- 

 quacy of classical small-deflection theory for predicting elastic-instability 

 pressures. Experimental studies conducted at the Model Basin have pro- 

 vided test data which substantiate these theoretical findings; see for 



example. Reference 26. A more detailed discussion of the elastic-insta- 



27 

 bility problem for ring- stiffened cylindrical shells is given by Reynolds, 



with some emphasis on large-deflection theory versus small-deflection 



theory for the hydrostatic pressure case. Another excellent discussion of 



28 



the instability problem has been given by Fung and Sechler. 



For the reasons emphasized above, no detailed consideration of the 

 large-deflection problem will be given in this presentation, so that our 



44 



