Multiscale Modeling of Defects in Solids
What makes defects form in the wing of an airplane, and how
do these grow and change until they eventually make the wing break? Until a few
years ago, scientists and engineers had to wait until a defect, or crack, was
big enough to be visible before they could predict how much longer the wing
could be safely used. But a multi-disciplinary team from Cornell University,
with support from the National Science Foundation, has devised a way to study
the evolution of these cracks from the time that they involved only a few
atoms. Their project was called "Multiscale Modeling of Defects in Solids."
The Cornell team consisted of Principal Investigator James
Sethna, from the Laboratory of Atomic and Solid State Physics; and co-PIs
Christopher Myers, from the Cornell Theory Center; Paul Dawson, from the
Mechanical and Aerospace Engineering Department; and Anthony Ingraffea, the
Dwight C. Baum Professor of Engineering. The group explored novel software
engineering techniques to support making computer models at multiple length and
They did this by developing a virtual laboratory, a software
infrastructure called the Digital Material, which they used to explore and
develop new theories and models of defect dynamics, and to test and validate
those models over a wide range of length scales, from the atomic to the
microscopic. They used two computer languages, one to write the code that did
all the work, and a second, high-level, language called Python, to steer the
applications of the code. They also used Design Patterns, a programming
technique that supports flexible program composition.
Until the early 1990s, experts focused on what happened from
the time a crack was large enough to be seen. However, according to Dr.
Ingraffea, an expert in aerospace structural engineering, "What we weren't able
to do was predict how long it would take from the time the airplane was first
put into service until the time the cracks were a millimeter long, and that's
most of the life. So we were doing a really good job of predicting 10 percent
of the life and a miserable job of predicting 90 percent of it."
Today that has changed. According to Dr. Sethna, "We try
to understand the defectsstarting from the atomswell enough to tell
you what they will do under various circumstances." Doing this involves more
than measuring one defect or crack. Researchers also need to know how the
defect is affected by its environment as well as how it affects its
environment. "The crack by being there changes the behavior of the material
near it," said Dr. Sethna. "Other defects might fly towards the crack or fly
away from it due to the crack's existence." This is why it was important for
the team to systematically measure a crack and extract all the information that
engineers might want to know.
This involved a change in approach for both the physicists
and the engineers. As Dr. Sethna explains, "Engineers by and large like to try
and understand the materials ignoring the atoms. And the physicists spend a lot
of time thinking about the atoms. Our approach is to try to extract the
information you need from the atomistic scale to communicate to the longer
The outcome has important applications not only in the
airline industry, but also in reusable launch vehicles used by NASA, and in
gears used in transportation from helicopters to submarines. It's also of use
in the fields of heavy industry and microelectronics. And ultimately, research
could yield results that allow metals to be designed from the atomic scale up
to be more resistant to the development of defects.
In addition to the principle investigators, the project also
involved faculty, senior researchers and research associates, postdocs, and
graduate students from Cornell's Laboratory of Atomic and Solid State Physics,
the Cornell Theory Center, and the Departments of Mechanical and Aerospace
Engineering, Materials Science and Engineering, Civil and Environmental
Engineering, and Theoretical and Applied Mechanics.
Back to Top of Page