Finding out what happens on the atomic scale while a material is being created, tested or utilized is extremely challenging.  Often simulation can help provide a detailed glimpse into the process in question.  However simulations of this kind are typically severely limited in the amount of time they can simulate.  Molecular dynamics simulation, in which the motion of atoms is simulated explicitly, can typically only achieve simulated times considerably shorter than a microsecond.   At Johns Hopkins research is underway to significantly extend the applicability of these methods.

One such example is in the simulation of nanoscale friction.  Materials researchers can use tools like the atomic force microscope (AFM), where an atomically sharp tip is dragged along a material surface, to characterize the properties of the surface.  The value of this information is limited, however, because we still know relatively little the physics that controls the, often jerky, motion of the tip.  Insight from simulation could be beneficial, but while tips in experiments move at one micron per second or slower the short clock on most simulations mean they can only simulate tips moving in excess of a meter per second, one million times faster.  It is not clear that the physics controlling such high-speed motion is the same as the behavior seen at experimentally attainable speeds.

At Johns Hopkins we are developing techniques known as hyperdynamics simulation, where we alter the forces between atoms in a controlled way so that the simulation clock runs faster.  By speeding up the simulation clock we can reach much longer times and much more reasonable sliding speeds.  These techniques potentially have application far beyond nanoscale friction to other important scientific and technological applications, for example in understanding the behavior of MEMS devices and in predicting materials failure.