The nation's nuclear weapons were originally designed to last for 20 to 25 years. Each year, the directors of the Department of Energy's three nuclear weapons labs must certify that the stockpile weapons will perform as designed. If the performance of an older weapon becomes questionable, lab scientists must decide how to replace its aging parts in order to restore its peak performance.

Aging weapon parts are a major topic for materials scientists Paul Dunn, Rusty Gray, Dave Teter, and Dan Thoma, who are working to extend the lifetimes of stockpile weapons in two of Los Alamos' highest-priority programs. The goal of one of these programs is to extend the lifetime of a 25-year-old weapon system to 75 years, which will guarantee its performance until at least 2042. Materials science is key to these efforts because, as Thoma says, "The only things that change in a stockpiled nuclear weapon are the materials." These scientists must determine how the materials are changing so they can predict when a part will fail.

No new warheads have been produced for over 10 years, a hiatus that is rapidly increasing the average age of stockpiled weapons.

Materials science has become more central to the weapons program because of a change in how weapons research is done. During the Cold War, researchers designed and built nuclear weapons without completely understanding the materials. They simply used the materials and fabrication processes that made a weapon explode at its nominal yield when tested at the Nevada Test Site. However, the end of the Cold War brought a ban on nuclear testing. Now, nuclear weapons are "tested" in computer simulations, which do require a full understanding of weapon materials, because a simulation must accurately predict how weapon parts will behave during the detonation.

However, there are many materials in a nuclear weapon, including metals, high explosives, polymers, and ceramics. The aging of any one of these materials could affect weapon performance, which is why scientists with a wide range of materials expertise are working to understand aging processes. The challenges these scientists face are illustrated by the aging issues for uranium, the specialty of Dunn, Gray, Teter, and Thoma.

Old Uranium
Uranium is the main heavy metal used in a weapon's "rad case" to redirect the x-ray radiation produced by the weapon's fission primary. It is also sometimes used in other weapon parts.

Uranium and its alloys age in several ways. Like steel, pure uranium "rusts" when exposed to the oxygen in air. It is also corroded by hydrogen. Although nuclear warheads are sealed in airtight metal containers to reduce oxidation and corrosion, the high explosives, plastics, and other organic materials also in the container emit tiny amounts of oxygen, hydrogen, and water vapor that, over time, can cause problems.

Uranium alloys also change their crystal structures, or phases, over time, which also presents aging problems. Materials scientists manufacture a part to have a specific phase in order to optimize its strength, density, or corrosion resistance. However, the strain a part accumulates during fabrication and the temperature variations a weapon experiences in the field can, over time, change the phase, thereby degrading a part's properties.

Subjected to the considerable heat given off by a weapon's radioactive plutonium, for example, a uranium part—and all other weapon parts—can reach temperatures as high as 40°C (about 100°F). A weapon can also experience temperature extremes in its storage environment, such as a desert. Temperature-induced phase changes that degrade uranium's mechanical properties are a major concern.

Predicting the Future
A logical first step in deciding when to replace a weapon's parts is to examine the parts. Each year, technicians randomly select ten or eleven stockpiled weapons for each weapon system, take them apart, and inspect, weigh, and x-ray their parts. Some of the parts are then sent to the weapons labs for further study. From an analysis of the parts, materials scientists must predict when they will fail. Predictions may have to extend many decades into the future.

For example, Thoma has developed a model to predict long-term changes in the ductility of a rad case. (Ductility measures a material's ability to inelastically deform without fracturing. It is important because rad cases become brittle with age.) Thoma has modeled the thermal migration of uranium atoms within the case material, which leads to phase changes that embrittle it. The model's predictions agree well with mechanical measurements on rad cases 2 to 9 years old. Thus, there is good reason to believe the model's predictions for rad cases many decades old. Extrapolating much further into the future, however, requires accelerated aging tests. For example, the aging of uranium alloys may be accelerated by heating them in the presence of various gases.

The effects of aging on a uranium rad case. The squares represent measurements of the mechanical strengths of rad cases that were pulled from the stockpile for inspection. Higher mechanical strength means that a rad case is more brittle. When the case becomes too brittle, it has reached the end of its life. The line represents the mechanical strength predicted by a theoretical model developed at Los Alamos. The close agreement between theory and measurement implies that the cases will not have to be replaced for many decades.

To predict a part's lifetime, scientists must also know how the part was made, because a material's properties strongly depend on how it is processed. For the older weapons, details on part fabrication now considered important were not always recorded. Often, however, scientists can retrieve the missing information by applying knowledge and techniques developed through decades of weapon materials research. Thoma points to this sort of detective work as an excellent example of how studies whose goals are to clarify everything possible about a material—without an immediate payoff—can help solve problems that arise later. This is the nature of science-based stockpile stewardship.

Replacing Old Parts
Once they have decided that a part must be replaced, materials scientists must decide how to manufacture its replacement, since many of the nation's weapons plants have been dismantled. For small numbers of parts, they may be able to use existing small-scale fabrication facilities.

For example, Los Alamos has several sophisticated small-scale facilities for processing uranium and plutonium parts, including uranium- and plutonium-casting facilities. The Lab also has the specialized expertise needed to work with these and other weapon materials and, in fact, has begun developing the capability to manufacture replacement pits, the main component of a fission primary.

However, a process used to produce replacement parts must be cost-effective and must comply with environmental regulations that did not exist during the Cold War. Further, because the material properties depend on processing steps, the parts produced by a new process may not be exact replacements. Although newer technology may produce better parts, the fact that the parts are different can lead to changes in their performance and aging.

Gray illustrates the problem with potato chips. The same ingredients are found in traditional chips sliced from potatoes and in the injection-molded chips that stack snugly inside a can. But the chips' properties—their appearances, textures, tensile and bending strengths, even their tastes—differ because of their different processing.

Dunn uses this analogy in discussing the replacement of rad cases. During the Cold War, rad cases were forged. To replace them now, it would be cheaper to cast new cases from the uranium melted down from old cases. However, casting produces a different microstructure in the uranium than forging does, which could produce compositional inhomogeneities that alter phase stability. Dunn and his co-workers are working to determine if such microstructural differences could have serious adverse effects. [figure: casting]

In addition, many materials—particularly the weapons' polymers—are no longer available, or the processes used to make them have changed so much that the materials' properties are quite different from those of the original polymers. Lab scientists must also find solutions to these problems.

It was, of course, much easier to determine the effects of a new fabrication process during the Cold War: a weapon containing parts made with the process was simply tested. But with weapon testing no longer an option, the complete effects of a new process can now be studied only with computer simulations.

Guaranteeing Shelf Life
The challenge of extending the lifetimes of nuclear weapons is complex and requires the coordination of many varied research efforts focused on how the properties of weapon materials change over time. These efforts include studying how aging and processing affect a material's microstructure and, in turn, how microstructure affects the material's bulk properties; extracting data on material properties under conditions that resemble those of a nuclear detonation; and accurately modeling the material properties in computer simulations of weapon performance.

Cost-effectively extending the shelf lives of weapons is critical to maintaining the nation's nuclear deterrent. It would cost about $1 billion to build a new facility to manufacture rad cases or other uranium replacement parts. A new facility for making plutonium parts would cost even more. These costs make it important to analyze all the options available for extending weapon shelf life.