2.8 Evaluate the test data

The next table shows the ambient virgin burst data together with the data from the cryogenically tested cylinders for comparison purposes. The ambient virgin burst data is the data obtained from hydrostatically bursting the cylinders at ambient conditions to establish baseline performance. The cryogenically tested cylinders are those cylinders that were filled with liquid nitrogen, pressurized to their designated test pressures three times, brought back to ambient temperatures and subsequently hydroburst.

HEI TEST DATA SUMMARY AMBIENT PERFORMANCE BASELINE VS. POST-CRYOGENIC TESTING PERFORMANCE

Overall performance of the pressure vessels is based on both the mean burst of each group of cylinders and also the coefficient of variation (CV). CV is a means of defining variability and is instrumental in designing a safe pressure vessel and in determining safety factors. Having a lower CV means having less variability in performance and being able to design a lighter weight pressure vessel.

The results of the testing expanded the overall database of cryogenic pressure vessel performance and demonstrated cryogenic capabilities of different material systems. Most of the material systems tested were degraded in overall performance as a result of the cryogenic testing. However, the cylinders made with T-1000 prepregged with UF3325-95 had a slightly higher mean burst and had a much smaller CV, increasing overall performance by about 10%. HEI is in the process of filling out a Disclosure of Invention and New Technology form for this new technology. The T-1000 with the NASA experimental resin showed no overall performance degradation. Its mean burst went down slightly, but its CV was decreased enough to make up for it. It should be noted that many of the CV's of the material systems decreased.

HEI also hydrostatically burst test the cascade cylinder for "fire for effect". This cylinder had essentially seen every cycle that all the other cylinders had seen combined (both pressure and temperature). The following chart shows the data obtained from both the virgin hydroburst test, which was conducted on a cascade cylinder to establish baseline performance, and the post cryogenically tested hydroburst data of the actual cascade cylinder used in the test assembly. It is interesting to note that the cascade cylinder degraded in performance by 15% after extensive cycling in both temperature and pressure.

CASCADE CYLINDER DATA

A non-linear finite element analysis was conducted on these cylinders to determine to what level of strain the cylinders had been taken during cryogenic testing. The following chart details the percentage of ultimate strain each set of cylinders was pressurized to.

STRAIN LEVELS OF EACH CONFIGURATION OF CYLINDERS

The next chart demonstrates the overall performance of the cryogenically tested cylinders. They were pressurized (to pressures established using the test logic defined in section 2.2) at cryogenic temperatures using a safety factor of 1.5 (assuming ambient conditions and no cryogenic testing).

The cylinders and test scenario were designed using ambient performance data and a targeted 1.5 safety factor. The actual test data, as summarized in the above chart, shows that as a direct result of the cryogenic environment the actual safety factor was significantly reduced for most of the materials tested. This would obviously affect the reliability of any system using such pressure vessels.

The data obtained from testing Dr. DeLay's cylinders follows.

NASA TEST DATA SUMMARY AMBIENT PERFORMANCE BASELINE VS. POST-CRYOGENIC TESTING PERFORMANCE

The results from the tests conducted on Dr. DeLay's cylinders suggest that the performance of these different configurations has been affected. The cylinder made with TRH50 prepregged with WDE-3D-1 was affected the most. The IM9, IM7, and T-1000 systems saw less degradation in performance. With the limited number of tests, it was impossible to establish a coefficient of variation (definition of variability) for the different configurations tested. More testing would need to be completed to get a better understanding of overall performance of the different configurations.

Several cylinders were sectioned and prepared for microscopic photographs in order to evaluate strain cracking. Previous data suggests that strain cracking due to cryogenic temperatures damages pressure vessels. The initial intent was to photomicrograph cylinders that had not been cryogenically tested and compare those to photomicrographs of cylinders that had been cryogenically tested. HEI hoped to be able to use intact fragments from cylinders burst at ambient conditions for the comparative photomicrographs (the following photomicrographs show just such cylinders that have been hydrostatically burst to failure, sectioned, and polished). Each photo shows cracks that go through the hoop material. These cracks are likely due to the overstressing of and indeed the failure of the fibers, making this an invalid comparison. The idea to make a comparison was therefore no longer valid. The cylinders made with Zylon fiber were nearly impossible to polish due to the abrasive nature of Zylon. Further explorations into polishing the Zylon samples were pursued no farther due to time and budget constraints.

Cracked Hoop Material of Hydro burst Cylinder	Cracked Hoop Material of Hydro burst Cylinder with Helical Material in View

Photomicrographs were also taken of cylinders that had been cryogenically tested, but not hydroburst. These photomicrographs follow. Notice that the T-1000 UF3325-95 sample, which configuration performed very well when hydroburst, demonstrates no major cracks in the composite material. The T-1000 Epon 828 sample, which performed poorly when hydroburst, shows one very large crack in the hoop material. The T-1000 Cryo Experimental Resin sample, which showed no degradation when hydroburst, also shows severe cracking in the hoop material.

Cryogenically Tested T-1000 UF 3325-95 Showing Few Cracks	Cryogenically Tested T-1000 Epon 828 With Severe Cracking Through Hoop Material
T-1000 Cryo Experimental Resin with Severe Cracking in Hoop Material

The results obtained from the photomicrographs are inconclusive. One cylinder that performed poorly had severe cracking, while another cylinder with severe cracking performed well. Further exploration would be necessary to better understand the cracking phenomenon and how it affects cylinder performance.

It is important to note that the cylinders built with TRH50 by Dr. DeLay, incorporated an experimental crack resisting barrier between layers. It was intended to reduce or even eliminate the cracking that occurs from cryogenic testing with this barrier. Although, in this case, the barrier did not increase burst performance of the cylinder, it is believed that a crack resisting layer applied utilizing an optimized process (yet to be fully determined) with optimized materials will increase performance of a cryogenic cylinder. HEI and Dr. DeLay are in the process of completing a Disclosure of Invention and New Technology form on this new technology.

Previous 1.0 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 4.0 5.0 Appendix Next

HyPerComp Engineering, Inc.; SBIR NAS8-03027 Final Report