PROCEEDINGS ARTICLE | April 10, 2014
Proc. SPIE. 9061, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2014
KEYWORDS: Thin films, MATLAB, Data modeling, Polymers, Composites, Resistance, Computer simulations, Nanocomposites, Single walled carbon nanotubes, Carbon nanotubes
Over the last few decades, carbon nanotube (CNT)-based thin films or nanocomposites have been widely investigated as
a multifunctional material. The proposed applications extend beyond sensing, ultra-strong coatings, biomedical grafts,
and energy harvesting, among others. In particular, thin films characterized by a percolated and random distribution of
CNTs within a flexible polymeric matrix have been shown to change its electrical properties in response to applied
strains. While a plethora of experimental work has been conducted, modeling their electromechanical response remains
challenging. Furthermore, their design and optimization require the derivation of accurate electromechanical models that
could predict thin film response to applied strains. Thus, the objective of this study is to implement a percolation-based
piezoresistive model that could explain the underlying mechanisms for strain sensing. First, a percolation-based model
with randomly distributed, straight CNTs was developed in MATLAB. Second, the number of CNTs within a unit area
was varied to explore its influence on percolation probability. Then, to understand how the film’s electrical properties
respond to strain, two different models were implemented. Both models calculated the geometrical response of the film
and CNTs due to applied uniaxial strains. The first model considered the fact that the electrical resistance of individual
CNTs changed depending solely on its length between junctions. The other model further explored the idea of
incorporating strain sensitivity of individual CNTs. The electromechanical responses and the strain sensitivities of the
two models were compared by calculating how their bulk resistance varied due to applied tensile and compressive strains.
The numerical model results were then qualitatively compared to experimental results reported in the literature.
