an 0-ring groove. Since the flange is usually welded to the steel pressure hull, 

 it would be cheaper to replace a damaged 0-ring groove in the viewport as 

 opposed to a damaged one in the flange. 



Structure 



With the design in Figure A-2 exhibiting advantages both operationally 

 and economically, it remains only a question of whether the viewport can 

 structurally withstand the stress concentration due to the presence of the 

 0-ring groove. The earlier seal designs in Figure A-1 did not structurally 

 affect the viewport. Designers probably did not use the design in Figure A-2 

 because of the unknown effects of the stress concentration; there was no 

 thorough stress analysis until Reference 5 was completed. 



The possibility does exist with 0-ring seal no. 2 in Figure A-1 , that 

 the softer acrylic will creep down into the 0-ring groove in the flange and 

 cause an extremely high stress concentration that could be catastrophic with 

 long term loading. 



The 0-ring groove in the viewport seal design of Figure A-2 was 

 located near the high-pressure face because the results discussed in the main 

 text of this report indicated that the stress distribution was nearly hydro- 

 static in this area. The rest of this appendix is concerned only with the 

 design in Figure A-2 and the structural response of the viewport to the 

 presence of the O-ring groove. 



SCOPE OF INVESTIGATION 



To thoroughly determine the structural effect of the groove on the 

 viewport, both an experimental and analytical phase were conducted on the 

 design shown in Figure A-2. 



Experimental Phase 



The viewport used to test the O-ring design was very similar to the 

 t/d = 0.45 viewports used in the main text and is shown in Figure A-3. The 

 temperature during the test was maintained between 65 and 75°F because 

 this would impose a more severe test than a lower temperature. While the 

 pressure was held at 4,000 psi, the displacements at the center of the low- 

 pressure face were recorded until the displacements became asymptotic 

 within the measurement accuracy. The machining procedure, pressure vessel 

 type, and testing procedures are all discussed in Appendix B. 



36 



