ABSTRACT 



The results of numerical calculations of the general-instability strength 

 of ring- stiffened circular cylinders are presented in graphical form. The calcu- 

 lations are based on Kendrick's "second solution" which is published in Naval 

 Construction Research Establishment Report No. 244 (Part III). The collapse 

 pressures from these graphs agree within 10 percent with those computed by 

 Kendrick's theory throughout the normal range of submarine geometries. 



INTRODUCTION 



One consideration in the structural design of submarine pressure hulls is the possi- 

 bility of collapse by general instability, i.e., large deformations of frames and shell between 

 holding bulkheads. It is also recognized that the elastic general-instability pressure governs 

 the extent to which imperfections reduce the load-carrying capacity of the frames. For this 

 reason, an accurate determination of this pressure can be useful in frame design even though 

 it is far greater than the pressure encountered at normal operating depth. 



The most reliable of several theoretical investigations of the general-instability prob- 

 lem appears to be the recent work of Kendrick conducted at the Naval Construction Research 

 Establishment, Rosyth, Scotland. This was published in the form of two separate analyses.''*^ 

 Both employ the same general approach, but Reference 2 (Kendrick, Part III) is a more general 

 treatment than Reference 1 (Kendrick, Part I) and always gives a lower collapse pressure. 



While Kendrick's analysis is a significant advance in the study of stiffened cylinders, 

 its application to submarine design is difficult because of the lengthy calculations required. 

 Recently Bryant, ■^ also o^ the Naval Construction Research Establishment, developed an 

 approximation which agrees closely with Kendrick's Part 1 solution. Because of its sim- 

 plicity, Bryant's formula is a valuable aid in design calculations. 



These solutions have been examined at the Taylor Model Basin, and several tests^ 

 have been conducted with machined models to provide experimental evaluation. In general, 

 agreement between experiment and theory was good, the failure pressures usually being slight- 

 ly lower than the prediction of Kendrick Part I or Bryant, and slightly higher than that of 

 Kendrick Part III.* The difference in the pressures given by these solutions is insignificant 

 for cylinders with light frames but becomes much larger for heavy frames such as used on sub- 

 marines. It was found, in fact, that for contemporary submarine geometries, Bryant's formula 



1 References are listed on page IS. 



* It should be noted that this analysis contains two solutions. The first represents a physically impossible 

 buckling configuration but is presented because it gives a lower collapse pressure. In this report, all mention 

 of Kendridc Part III will be confined to the second solution of that analysis. 



