Filament Wound Pressure Vessels & Space Shuttle Disasters

One of the most visible symbols of American ingenuity and exploration over the past 50 years has been the progress and perseverance of the National Aeronautics and Space Administration (NASA). It is the epitome of human curiosity. It is a constant example of how we live with unknowns for only as long as technology limits our ability to explore; and sometimes we try and explore those unknowns before technology is ready.

A total of 18 astronauts (12 of which were Americans) have perished while on space flight missions, and more have lost their lives while preparing or testing for them. Regardless of how bright your engineering team or how much money you have, complex systems often find their demise at the hand of a simple problem.

In 2003, the Columbia Space Shuttle disintegrated on re-entry, killing 7 astronauts aboard. The accident was caused by a piece of foam insulation that broke off from the Space Shuttle external tank and struck the Orbiter’s left wing.

After the accident, an investigation was launched into the Culture of Safety which existed at NASA. The results were disturbing. 2005 estimates for the probability of a catastrophic failure over the life of an Orbiter (estimated 100 missions) were one-in-five. The investigation put it succinctly:

“Organizations that deal with high-risk operations must always have a healthy fear of failure – operations must be proved safe, rather than the other way around. NASA inverted this burden of proof.” Thus, change in philosophy was mandated: the design and its operations must be proven to be safe.

One of the studies commissioned after this report was the analysis of the Orbiter’s composite-wrapped pressure vessels. Each Orbiter vehicle contains 24 Kevlar-COPVs for pressurized helium and nitrogen at pressures up to 5,000 psi.

Starting in the 1960’s and 70’s, the potential for Kevlar reinforced vessels was recognized at the NASA Lewis Research Center. Early research into filament winding reported weight savings of 25%-30% over comparable all-metallic vessels. Today, filament wound pressure vessels are essential to numerous NASA power and environmental systems. The majority of older vessel overwraps are made of Kevlar®-49/Epoxy Composites while the newer vessels have Carbon/Epoxy overwraps. Incorporating COPVs instead of traditional all-metal tanks reduced the Orbiter weight by 752 pounds.

Early on, the primary failure mode of the vessel was considered to be fatigue of the thin metal liner, either in the parent material or in the welds. The design requirement for the liner was Leak Before Burst (LBB). By definition, LBB requires that material defects or fatigue cracks in the liner manifest themselves in such a way that if a crack grows and penetrates through the liner, it would continue growing slowly enough for the gas to slowly escape.

However, from baseline tests, the following preliminary findings were released :

  • Kevlar/Epoxy (K/E) COPV are susceptible to the stress-rupture failure mode that can result in an explosive release of stored energy,
  • Risk is a strong function of fiber stress level, time at pressure (adjusted for temperature),
  • Likelihood of stress rupture failure in Orbiter COPV is significantly higher than previous predictions, and
  • Reliability models and supporting database need to be updated to reflect new knowledge as it is gained
  • Given the new NASA safety regulations, COPVs underwent extensive physical testing.

The original program requirements were for 100 missions in 10 years. Cycle testing was performed on test articles to show that each vessel liner could survive a factor of 4 on what was anticipated for 100 missions. The results of the study are beyond the scope of this post, however, twenty tanks were experimentally cycled in fatigue. For the most part, LBB was observed, however compressive (buckling) failures occasionally occurred but were deemed to be the result of unrealistic conditions. Several important conclusions were drawn from their work and if you are interested in composite pressure vessels, they are worth a read.

*Bonus 1:

When all design and maintenance costs are taken into account, the final cost of the Space Shuttle program, averaged over all missions and adjusted for inflation, was estimated to come out to $1.5 billion per launch, or $27,000 per pound.. This should be contrasted with the originally envisioned costs of approximately $657 per pound adjusting for inflation to 2013. By comparison, Russian Proton expendable cargo launchers (Atlas V rocket counterpart), are said to cost as little as $110 million (approximately $2,300 per pound).

Bonus 2:

Most of the smoke billowing up from a space shuttle launch is not exhaust. It is water vapor from the pool of water under the shuttle designed to absorb the acoustic shock waves that could otherwise tear the shuttle apart.