2/13/02 - HC Updated 1/22/02 Updated 1/29/02 Creep measurements on carbon fiber The material is .507" wide, 10" long, and .040" thick. The bulk modulus as estimated by the manufacturer data is about 9 mpsi. Data is taken in a clean room where temperature is maintained at 70F and the relative humidy is 30% 12/16/01 loaded with 106 lbs initial reading before loading = 0.018" after loading =0.0074" 5 hrs later =0.0070" 24 hrs later =0.0069" 92 hrs later =0.0068" 700 hrs later =0.0066" 1/9/02 Loaded with an additional 19.9 lbs Intital reading =0.0066" after loading =0.0051" dial indicator may have been bumped 16 hrs later =0.0049" 137 hrs later =0.0048" 167 hrs later =0.0048" 1/15/02 Removed the 19.9 lbs Initial reading =0.0064" 24 hrs later =0.0065" 144 hrs later =0.0065" 1/23/02 Added the 19.9 lbs Initial reading =0.005" dial indicator may have been bumped 1 hrs later =0.0049" 55 hrs later =0.0048" 1/25/02 Removed the 19.9 lbs Initial Reading =0.0053" dial indicator may have been bumped 72 hrs later =0.0064" 84 hrs later =0.0064" 1/29/02 Added the 19.9 lbs Initial Reading =0.0050" dial indicator may have been bumped 24 hrs later =0.0050" 120 hrs later =0.0049" 192 hrs later =0.0048" 2/7/02 Removed the 19.9 lbs Initial Reading =0.0063" dial indicator may have been bumped 72 hrs later =0.0063" 84 hrs later =0.0063" 2/12/02 Added the 19.9 lbs Initial Reading =0.0050" dial indicator may have been bumped Comments Intitial tensile stress ~5225 psi, expected stress on window is ~2-3,000psi Initial loading of the window will cause the window to bow out about 1 inch. Expected creep on a 160" sample with 5KSI loading = .0128" Every .010" of linear stretch corresponds to an additional .25" bowing of the window in the horizontal direction. This initial amount of creep is acceptable. Life of the window. If every 24 hours the barometric pressure changes .07 psi (gonzalo.er.anl.gov/ANLET/text)- which inturn changes the volume of the tank by 10ft^3, which changes the static head on the tank due to a height change in the expansion bellows by .2 feet. The nominal static head on the window is about 18 ft so this change is about a 1% change in loading per day. from the testing we know that a 18% change in loading causes about 0.0001" of permanent creep in the window. Approximating a 1% change is 0.0000055" of creep per day. The max allowable creep is about 0.010" which indicates a window life of 1800 days or 5 years. Calculating the life of the window based on the actual daily pressure changes recorded for a 3 month period gives a life of 2.5 years. The life of the window would also be reduced by daily temperature changes. The steel window frame will expand and contract due to thermal changes, however the carbon fiber will not. This extra stress cycling has not been calculated. An estimate is 160" * 5e-6 *10F=0.008" stretch. The stress would be approximated by 10e6*.008/(160)=500 psi or about 25% the total stress. To reduce the amount of creep- the window thickness could be increased to reduce stresses. Increasing the life by a factor of 4 would require an estimated increase in thickness of 2 times or 0.080". The radiation length would be .87% of a radiation length. The initial bow of the window would be less, which would also increase the creep life of the window. Comparing the ease of manufacturing, cost, and radiation length of a thicker carbon fiber window may drive the window material to aluminum. The radiation length of .048" aluminum is 1.4% Further testing to be done... continue current creep test with cycling the loading, Creep on aluminum, creep on a thicker carbon fiber.