A stochastic approach towards a predictive model on charge transport properties in carbon nanotube composites

Abstract A probabilistic approach is taken towards modeling the electrical transport properties of carbon nanotube composite materials. A pseudo-random generation method is presented with the ability to generate 3-D samples with a variety of different configurations. Periodic boundary conditions are employed in the directions perpendicular to the transport to minimize edge effects. Simulations produce values for drift velocity, carrier mobility, and conductivity in samples that account for geometrical features resembling those found in the lab. All results show an excellent agreement to the well-known power law characteristic of percolation processes, which is used to compare across simulations. One of the main objectives of this study was to determine optimum simulation box-sizes that minimize size-effects without rendering the simulation unaffordable. The effect of sample morphology, like nanotube waviness and aspect ratio, on charge transport within CNT composites is also evaluated. The accurate prediction of conductive properties as a function of the physical characteristics of the composite material, will aid in the design of composites that can, for instance actively monitor structural changes due to compression, stretching, or even damage through their effect on the conductivity.