Text Size
The NESC Assesses the Reliability of Composite Overwrapped Pressure Vessels (COPVs)
March 29, 2012

The NESC team will be testing composite strands and developing analytical methods to improve confidence in reliability estimates.  Collection of strand data from testing can provide a more comprehensive understanding of COPV stress rupture behavior.  After the assessment is completed, reliability models of the stress rupture failure mode of COPVs will be available to help ensure safety and reliability of COPVs.[image-62]
NASA COPVs are commonly used for gas and propellant storage in spacecraft and launch vehicles.  Because most COPVs are used to store high-pressure gas, the consequence of a COPV rupture can be catastrophic to surrounding spacecraft structure and components, equivalent in energy to the release of several pounds of trinitrotoluene (TNT) depending on the size of the COPV.  The safe and reliable use of COPVs is dependent on preventing any rupture failures.  Various rupture failure modes of COPVs are addressed during engineering design, but the longer-term stress rupture failure mode is not well understood.  To date, there have been no failures of COPVs in flight or in flight systems on the ground that can be attributed to stress rupture, but failures have been observed during testing at elevated stress conditions.  The relationship between these observed failures during test and flight vessel usage is not well understood, but the result of a COPV failure due to stress rupture would most likely lead to the loss of the spacecraft.  Because the consequences of a stress rupture failure are significant, better understanding is needed.

Stress rupture is a function of time and tensile stress in the overwrap.  The risk of a COPV failing in stress rupture has been predicted in the past based on empirical models because the stress rupture mechanism is not known.  These models use established statistical techniques.  Data are available in the literature for high stress levels, where the time to failure is relatively short.  However, it is difficult to acquire the large volume of data needed for the long periods of time required to generate failures at lower stress levels.  Thus, probabilistic models are necessary to extrapolate the data to predict reliability at the lower stress levels used for flight COPVs.  For example, Figure 2 shows raw load versus time data taken on IM6 carbon composite strands.  Data were obtained at four different stress levels.  The plot shows the large variability of the time to failure within each stress level.  Figure 2 also shows that the probability of a sample surviving the test was a function of stress level.  Additional modeling was also required to establish the relationship between the strand test results and flight COPVs.[image-47]In two recent NESC investigations into stress rupture, "Shelf Life Phenomenon and Stress Rupture Life of Carbon/Epoxy COPVs" and "Orbiter Kevlar/Epoxy COPV Flight Rationale Technical Assessment Report," attempts to quantify the reliability of COPVs for the International Space Station (ISS) and Space Shuttle Programs were made based on the existing data and modeling techniques.  While these investigations improved understanding of existing data, questions remain regarding the quality of the data on which they are based and the subsequent probabilistic models used to provide reliability estimates.  Because of the lack of applicable data and inconsistencies in existing data sets, the applicability of the existing models to ISS and other future NASA flight COPVs is unclear.  As a result, models based on existing data are not considered sufficiently reliable to address this failure mode.[image-69]

To improve confidence in reliability estimates, the NESC team will be testing large quantities of composite strands fabricated from the same materials as those being tested in a COPV stress rupture test that are being conducted by the ISS.  Collection of strand data can provide a more comprehensive understanding of COPV stress rupture behavior because larger quantities of strands can be tested at a lower specimen cost to increase the overall stress rupture database.  Strand testing will be conducted at the White Sands Test Facility.  The analytical skill to resolve the complex probabilistic modeling required to extend test results to NASA flight COPVs will be provided by a team of engineers and statisticians.[image-4]

After completing the testing and analytical work, reliability models of the stress rupture failure mode of COPVs will be available to ensure continued safe and reliable use of COPVs throughout NASA.


Sample Stress Rupture Area
Figure 2. Sample Stress Rupture Data The figure shows failures as a function of time.
Image Token: 
Typical NASA Flight COPV
Figure 1. Typical COPV
Image Token: 
Examination of Strand Test Stands
Examination of Strand Test Stands designed by Lawrence Livermore National Laboratory.
Image Token: 
Close-up of test strand.
Close-up of test strand.
Image Token: 
Page Last Updated: September 18th, 2013
Page Editor: NASA Administrator