Nanotechnology deals with investigating very small physical systems, about 50000 times smaller than the diameter of human hair. Researchers have observed that, in the nanoscale, physical behavior depends on the system size. Scientists around the world have constantly tried to utilise this behavior in developing new devices for improving the life on earth. For example, there has been numerous investigations on the physical behavior of fluids confined in nanoscale channels like carbon nanotubes and their potential applications in water filtration, energy harvesting, drug delivery etc.
A major roadblock in the rapid development of this field is the lack of accurate theoretical models to predict the changes in the dynamical behavior of physical systems with respect to the domain size. If we look at computational investigations in this field, molecular dynamics simulations have been in the forefront in developing a theoretical understanding of the nanoscale systems. But, such investigations are limited by issues like, the lack of computational power, or lack of accurate force fields to model the size effected behavior etc. Recent works in the literature about data-driven discovery of partial differential equations, have demonstrated the use of machine learning algorithms as a competent tool to develop theoretical models for physical systems. In this work we try to explore the possibilities of using machine learning strategies to understand the behavior of nanoscale systems and develop physical models for the dynamical processes occuring at this scale.
Department of Mechanical Engineering, Amritapuri
It is ideal but not mandatory for the applicant to have a basic understanding of programming languages like python.
Asst. Professor, Mechanical Engineering, School of Engineering, Amritapuri