Figure 2. Cross section of a typical viewport. 



The six viewports were tested simultaneously in three refrigerated 

 pressure vessels. Displacements and post-test observations served as the 

 recordable data. Details on the equipment and test procedure used are in 

 Appendix B. 



Inasmuch as an undersea vehicle might make many dives, cycling of 

 the pressure load was included as part of the test procedure. The simulation 

 of actual dives required simultaneous changes in pressure and temperature. 

 As the pressures were changed from to 4,000 psi, the temperature simul- 

 taneously was changed from 70 to 35°F. It was believed that the most severe 

 test of the viewports would be to cycle the maximum creep strains occurring 

 in each viewport under the 4,000 psi load. Therefore, the period of the 

 cycle for all viewports was determined by loading the least conservative 

 design, t/d = 0.45, and letting the viewport displace until the displacement- 

 time curve was asymptotic within the accuracy of the measuring system. 

 This resulted in a "load-on" time period of 500 hours with a "load-off" or 

 relaxation period of 100 hours; the time necessary to completely relax all 

 the strains so the viewport would return to zero displacement. The tests were 

 conducted over a 1-year period which resulted in 13 total cycles. 



