Over the past year, the two makers of large-body commercial airliners, Boeing and Airbus, have each experienced sudden and unanticipated problems with their premier new airplanes. Boeing’s 787 Dreamliner was temporarily grounded while the company worked out a fix to the overheating of the airplane’s lithium-ion batteries. Airbus has had a much more serious problem with its new super-sized A380. Although they have only been in service a short while, some A380s are beginning to suffer cracks in the brackets that hold the wings to the fuselage.
Both companies spent an enormous amount of manpower, money and time testing their aircraft. Simulations and tests were run not only on the whole plane or even major systems but on assemblies, brackets and other parts. Components and materials were subjected to detailed laboratory and engineering tests. Every time something was changed, such as the A380’s wing mounts, the pieces were tested extensively. Both planes went through exhaustive computer simulations. Finely the two companies logged literally thousands of flight hours under varying conditions to prove that the Dreamliner and A380 were safe to fly.
Yet, things went wrong. They always do. As a senior official of the European agency that certified the A380 stated “It’s impossible to test for everything – it’s simply too complex.” In part, it proved impossible to accurately simulate or replicate the extremes of real world conditions that the A380 confronted. Lithium-ion batteries are not exactly new technology. Yet, when they were integrated with all the other parts in the Dreamliner and flown under real world conditions, something unexpected happened. Both aircraft manufacturers have fixed these problems and the future for both aircraft looks bright.
The lesson to be drawn from the Dreamliner and A380 experiences is that it is impossible to rely solely on computer models and simulations when designing and building extremely complex systems. However, since the mid-1990s, the United States has relied on such modeling and simulation to ensure the safety and reliability of its nuclear weapons stockpile. The Stockpile Stewardship Program, in place since around 1999, professes to be able to guarantee the performance of U.S. nuclear weapons and their components. Because we have not tested a nuclear weapon in more than twenty years there is no way of knowing for sure how one will actually perform, particularly one that has been refurbished with new parts. Over time, nuclear radiation does funny things to materials. Yes, parts are tested in the laboratory. The behavior of the weapon is subjected to very elaborate computer modeling. But it is not the same thing as actually testing a weapon.
If we can’t always accurately predict the behavior of materials and systems such as lithium-ion batteries or aluminum brackets that are used every day, what does that say about the confidence we can have in the operations of a system as complex as a nuclear weapon. The Obama Administration is considering another round of deep reductions in the U.S. nuclear arsenal, including the elimination of most non-deployed weapons in the stockpile. If we cannot accurately predict the behavior of these systems but must rely on them for deterrence, what sense is there in reducing the numbers? Ironically, when the quality of the system is suspect, the quantity available may need to be greater than otherwise required.
Find Archived Articles: