Modeling the effect of uniaxial deformation on electrical conductivity for composite materials with extreme filler segregation

Abstract

In this work, the correlation between electrical conductivity and uniaxial deformation of a material with highly segregated distribution of conductive filler is studied. Multi-walled carbon nanotubes are used as a model filler. A numerical model that can be used to predict changes in conductive microstructure made of multi-walled carbon nanotubes in response to uniaxial deformation of material is proposed. The model takes into account the ability of nanotubes to assume various conformations and orientations during deformation. Numerical simulations are conducted for uniformly distributed multi-walled carbon nanotubes providing confinement of the filler in a two-dimensional film structure with high volume fraction of the filler. The embedded element method to conduct realistic and computationally efficient simulation of multi-walled carbon nanotube behavior during deformation of the composite material is implemented. Finally, the results of numerical simulations of changes in electrical conductivity of composite during deformation are compared with the experimental data to prove the correctness of assumptions used in the model.

Multiscale Numerical Modeling for Prediction of Piezoresistive Effect for Polymer Composites with a Highly Segregated Structure

“…First, the initial points (“seeds”) of the MWCNTs were distributed uniformly in a volume of 0.1 μm × 5 μm × 5 μm. The number of MWCNTs was chosen according to the volume fraction of the filler in the conductive layer (55%), obtained experimentally at the previous stage of study [ 36 ]. The layer thickness of 100 nm was calculated as the thickness of a conductive layer enveloping UHMWPE grains in solid-state processed UHMWPE/MWCNT nanocomposites with ~1% total volume fraction of the filler in the matrix.…”

“…First, the reason for not using the representative volume element as obtained by truncation of the initial volume to the boundaries of the “inner” volume was because, in that case, some MWCNT segments cut by the new boundaries might have lengths shorter than the statistical length. The presence of segments of different lengths can result in unpredictable outcomes [ 36 ] and, therefore, is undesirable. Second, periodic boundary geometry could not be created for the truncated volume due to the specifics of the meshing method selected in this work—the embedded element method—that eliminates all degrees of freedom of MWCNT elements.…”

“…Second, periodic boundary geometry could not be created for the truncated volume due to the specifics of the meshing method selected in this work—the embedded element method—that eliminates all degrees of freedom of MWCNT elements. Discussion on the advantages and disadvantages of the embedded element method applied to similar systems can be found in our publications [ 35 , 36 ].…”

“…To increase the computational efficiency of FEM simulations, the embedded element method for mesh creation can be used [ 32 , 35 , 36 , 37 ]. This method implies the creation of two separate independent meshes—for the matrix and for the filler—without the necessity to generate an identical node network at the interface.…”

“…At the previous stage of this work, the piezoresistive response of a thin MWCNT layer deposited on a polymer surface was investigated experimentally and numerically [ 36 ]. A good correspondence between numerical and experimental results was observed.…”