for relatively thick shells, e.g., R^/R^ =0.9 and h/R - i/9.5, (where R- 

 is the radius to the inside surface of the cylinder, Rq is that to the outside 

 surface, R is that to the median surface, and h is the thickness of the 

 cylinder). However, they found that these shell theories do not adequately 

 predict the stresses and radial deflections in the neighborhood of concen- 

 trated loads. At these locations, the radial deflections are predicted quite 

 accurately by the transverse shear deformation shell theories. Interest- 

 ingly enough, shear deformation theory was found not to improve the 

 accuracy of the axial displacements and the stresses. 



For the particular problem of a long circular cylindrical shell of 

 constant thickness subjected to a radial line load considered by Klosner 

 and Kempner, the authors found that the classical shell theories predict 

 deflections which are 8 percent smaller than those obtained from the 

 elasticity solution. Thus, if one would consider a ring- stiffened cylindrical 

 shell subjected to a uniform pressure, then the mcLximum error in the 

 calculated interaction load may be as great as 8 percent and will occur 

 when the spacing of the reinforcing rings is large and the rings are rigid. 

 As the rigidity of the ring frames decreases, the error decreases. Thus, 

 for ordinary pressure-hull design, the interaction loads are not significant- 

 ly different whether use is made of classical shell theories or three- 

 dimensional elasticity or shear deformation shell theories. What is 

 significant, however, is the difference in the stress distribution, and it is 

 with this consideration in mind that future work pertaining to pressure-hull 



119 



