Industrial materials must be "tuned" to meet special requirements.
This tuning usually involves adding or controlling defects,
through an extensive and expensive "trial and error" process.
For example, nickel alloys can be optimized by adding other elements
so that as-cast turbine blades are stronger at the high operating
temperature of a jet turbine engine. Past experience, as in the case of blade
alloys of a 10-15 year development process costing $50 M, discourages
many industries from researching new materials.
Theoretical simulations promise to speed this tuning process
by quickly identifying a promising experimental path.
The picture displays the effect of adding two extra silicon atoms
to a perfect silicon crystal. That there
are two extra atoms can be seen by mentally removing two of the the
three red atoms forming the central triangle; the remaining red atoms
relax to a plane aligned with the green plane underneath. This
triangular defect moves easily through the crystal, forming
even larger defects and altering properties. As the semiconductor
industry strives to produce smaller and hence faster computer chips,
control of these defects at the atomic level is increasingly important.
Our theoretical effort is to quickly identify the triangular and other
defects and understand their effect on properties. The link at the
bottom shows three triangle defects that were selected in the course of
a 100,000 time-step stimulation.
Simulations of which this illustration is just the beginning
can winnow the experimental route to tuning a material for a specific use.
Some analysis of paragraph: