from the original flat plating of Models OV-1 and OV-2, respectively. 

 These higher yield strengths are reflected in the 13- and 18-percent dif- 

 ferences between observed collapse pressure and the p^ pressure for each 

 of the models. Thus, the optimistic collapse strengths realized from the 

 tests of both of these models can be attributed to the beneficial effects of 

 strain hardening of the shell material. 



When the observed pressures of Models OV-1 and OV-Z are extrap- 

 olated to a common yield strength of 120, 000 psi, as shown in the last 

 column of Table 2, OV-2 has a higher collapse pressure. This can be at- 

 tributed to the more "balanced state of stress" for this configuration; see 

 Figure 7 . 



Table 2 also lists the "upper bound" pressures p^ for each of the 

 machined titanium models. The almost exact agreement between p. and 

 the observed collapse pressure in the case of Models OV-3 and OV-5 is 

 fortuitous. However, it indicates the high degree of efficient utilization of 

 the material that can be realized with sandwich construction. For Model 

 OV-3, a comparison can be made with the counterpart steel sandwich mod- 

 el (OV-l) where an appreciable difference between observation and "upper 

 bound" prediction can be attributed to the fact that the steel structure was 

 stressed into the strain-hardening range. The effects of strain hardening 

 were not realized with the titanium sandwich model (OV-3), even though 

 the geometric disposition of the material was the same in both cases. 



If a linear "correction" of the experimental collapse pressure for 

 the short titanium model (OV-3) is made to allow for the higher yield 

 strength of the shell nnaterial as connpared to the minimum specified, a 

 collapse pressure of 1 1 , 040 psi is realized. This value is indicative of 

 the maximum collapse strength which can possibly be realized with the 

 sandwich configuration shown in Figure 1, but it does not reflect the influ- 

 ence of residual stresses which may arise due to rolling and welding in 

 a fabricated prototype. These considerations would have a tendency to 

 "round off" the stress -strain curve even more so than that shown in Fig- 

 ure 20. Also, coupled with the destabilizing influence of a long compart- 

 ment length, they could lead to a value of collapse strength below 11, 040 

 psi. The question as to how much lower this strength wo\iLd be and whether 

 it would fall below the minimum design collapse pressure of 10, 000 psi, 



27 



