The influence of window fit on the critical pressure was investigated with 

 windows having the same D^/D; ratio and thickness (Figures A-3a and A-3b) fitted 

 into flanges with different major diameters (Figures A-la and A-lb). in one of the 

 flanges (Figure A-la) the pretest radial clearance between the windows and the 

 flange was either 0.001 inch (Figure A-3b) or 0.025 inch (Figure A-3a), while in the 

 other flange (Figure A-lb) the clearance was 0.125 inch. A radial clearance of 

 0.001 or 0.025 inch when the window is subjected to hydrostatic pressures above 

 10,000 psi is calculated to result in an interference fit between the window and the 

 flange, thus resulting in a lateral constraint of the window. The flange (Figure A-lb) 

 and window (Figure A-3a) assembly with the initially larger radial clearance of 

 0.150 inch, even when subjected to hydrostatic pressures that destroyed the window, 

 did not cause it to be wedged inside the flange opening. With such an arrangement 

 it was possible to determine whether the wedging in of the window in the flange 

 under hydrostatic pressure had any measurable influence on the critical pressure of 

 flat windows. 



DISCUSSION 



Relationship Between Critical Pressure and D^/Dj Ratio 



Tests to determine the relationship between critical pressure and Dq/Dj ratio 

 were conducted with five 2-inch (D^) windows in a 1.5-inch (D;) flange and five 

 4-inch (Dq) windows in a 1.5-inch (D|) flange. The windows were sealed in the 

 flange with the aid of silicone grease, which was liberally applied to the bearing 

 as well as the radial surfaces of the flat circular window. For both the 2-inch and 

 the 4-inch (D ) windows, the radial clearance between the window and the flange 

 was 0.025 inch. 



When tested to destruction, the average critical pressure of 2-inch (D ) 

 windows was 18,490 psi (Table C-1), while the critical pressure of 4-inch (D ) 

 windows was 19,190 psi (Table C-16). The small difference between the average 

 critical pressures of the 2-inch and the 4-inch (Dq) windows with a 0.5 t/Dj ratio 

 seemed to indicate that varying the Dq/Dj ratio from 1.33 to 2.67 did not signifi- 

 cantly influence the critical pressure of flat acrylic windows, since the maximum 

 collapse pressure found in 2-inch (D^) windows (18,900 psi. Table C-1) overlapped 

 the minimum collapse pressure found in 4-inch (D ) windows (18,800 psi. Table C-16). 



Since the critical pressures of windows with 1.33 and 2.67 D /D; ratios are 

 approximately the same so long as their t/D| ratios are identical, a flange with an 

 intermediate Dq/D; ratio of 1.5 was selected for the conduct of the main flat- 

 window study program. 



24 



