Long-Term Loading of Spheres 



Of the original eighteen spheres in 

 the ocean, three spheres have imploded 

 at depth. The time-to-failure for each of 

 the spheres has had to be estimated 

 because the clocks contained in two of 

 them have not been found. Sphere no. 3 

 evidently imploded on descent to the 

 seafloor at ^1,330 feet (1,320 m) because 

 the concrete fragments v^^ere widely 

 scattered. Spheres no. 1 and 7 had a 

 time-to-failure that was between the 

 beginning of the test and the day of their 

 first inspection. For these specimens the 

 concrete fragments were in a localized 

 area. 



Detailed data for these spheres and 

 the other spheres still intact on the 

 seafloor are given in Table 6. This table 

 also shows the calculations for the rela- 

 tive load level, i.e., the ratio of 

 sustained pressure to the predicted 

 implosion pressure, Ps/Pim > fo"" each 

 sphere. The relative load level for the 

 spheres changed between the beginning 

 of the test and the 5.3-year period as the 

 concrete became stronger with time. 

 Using data from Figure 7 for dry con- 

 crete and Figure 8 for saturated 

 concrete, the 28-day fog cure compres- 

 sive strength was adjusted to estimate 

 the strength at 5.6 years. 



The long-term loading failure 

 behavior for the ocean spheres is shown 

 in Figure 22. Data from spheres tested in 

 a pressure vessel (Ref 5) are also pre- 

 sented in Figure 22 and summarized in 

 Table 7. The results of Stockl (Ref 9) 

 are shown by an average data curve 

 which represents hundreds of tests on 

 uniaxially loaded specimens. With two 

 exceptions, the sphere results are in 

 agreement with Stockl's findings. From 

 his data, it appears that the safe relative 

 load level is about 0.75. 



Stockl presented an interesting dis- 

 cussion on long-term loading. When a 

 structure is initially placed under 



sustained load, the creep behavior of 

 concrete causes degradation of strength 

 by forming microcracks. At the same 

 time, in the presence of moisture, hydra- 

 tion of cement causes a strengthening of 

 the concrete. Long-term loading to 

 failure occurs when microcracks from 

 creep progress at a faster rate than 

 strengthening from cement hydration. 

 Since creep effects slow down with time, 

 strength gain from cement hydration has 

 a chance to catch up and overtake the 

 creep-induced strength reductions. The 

 rate of cement hydration is dependent 

 upon the amount of free cement 

 available within the concrete before it is 

 placed under sustained load. The 

 practical significance of this condition is 

 that once the "critical duration of load" 

 is over and hydration strengthening 

 occurs at a faster rate than creep 

 damage, the specimen will never fail 

 from creep. The critical duration of 

 sustained load can be days when applied 

 to young concrete and years for old 

 concrete. For the spheres in the ocean, 

 the critical duration of load was probably 

 on the order of months. Hence, according 

 to Stockl, any sphere that has not failed 

 after 5.3 years under sustained load 

 should not fail in the future. 



In Figure 22, for Sphere no. 1 {^,075 

 feet) and Sphere no. 7 (3,725 feet) the 

 time-to-failure is shown as a dashed line. 

 The datum point for the spheres was 

 plotted at 10 days as a conservative 

 estimate. For Sphere no. 3 (4,330 feet), a 

 wide scatter of fragments on the 

 seafloor indicated that the specimen 

 failed during descent. 



A valuable supplement to the failure 

 data was the results from the spheres 

 that had not failed. Figure 23 shows the 

 relative load levels for the spheres in the 

 ocean that are still intact. Stockl's curve 

 is shown to lie above the data. The 

 relative load levels experienced by the 

 spheres in the ocean are shown to 

 decrease with time as the strength of 

 concrete increases. 



26 



