Summary: HPC issues in the modelling of nanotubes
The application of Newton's equations of motion with suitably reliable potentials to
atomic systems is known as Molecular Dynamics. It enables modelling, e.g. of molec-
ular flow within nanotubes and of nanotube vibrations, with and without additional
adsorbed particles. The former has applications to Chemical Engineering issues
concerning membranes that can seperate, e.g. methane and hydrogen, the latter to
nanotechnology, specifically to the search for the ultimate resolution in mass sensitivity.
HPC aspects of such modeling that interest my group at the Technion include:
· Visualization, preferably three dimensional - we are porting the impressive 3D
implementation of our Technion AViz visualization code (made by Herwig Zilken
of Juelich) to the Technion as one focus of our current Umbrella project.
· Algorithms for efficient (preferably parallel) Molecular Dynamics simulations for
the relatively long ranged potentials needed to reliably respresent nanotubes. The
parallelization is in progress at the Technion, (in collaboration with Yuval Yaish,
Polina Pine and Ofer Filiba) but a stronger machine than our NANCO is needed
to bring this project beyond its development stage. The aim is to be able to study
nanotubes of experimentally realisable lengths.
· Protocols for handling and analysing the large amounts of data needed in the nan-
otube vibration project.
· Extensions to other nanotube systems such as polymer tubes - the second focus