Conductive CNT-polymer nanocomposites digital twins for self-diagnostic structures: Sensitivity to CNT parameters.

[Display omitted] • The digital twin of a CNT-polymer nanocomposite is developed. • CNT intrinsic resistivity cannot be neglected. • Uniform boundary potentials overestimate effective electrical conductivity. • Control of CNT torsion removes the dependence on the segmentation length. • The critical percolation index lies in the range 1.2–1.7. The paper describes digital twins for an electrically conductive isotropic CNT-polymer nanocomposite, using representative volume element with boundary conditions being periodic geometrically and electrically. During the CNT generation, the torsion and curvature of CNTs are controlled. The percolation threshold is analysed with scaling towards the infinite RVE. The digital twin sensitivity to model parameters is found as follows: (1) restrictions on the curvature/torsion ensure the independence of the homogenised conductivity on the CNTs discretisation; (2) the conductivity of the nanocomposite increases (for the same CNT volume fraction) with the increase of the CNT length and maximum torsion and decrease of the maximum curvature; (3) the percolation threshold, is in the range of 0.3–0.7% for different CNT parameters; it reduces in half with doubling the CNT length; the curvature and torsion have lesser influence. The critical percolation index is 1.2–1.7, which corresponds to the theoretical values for 3D networks. [ABSTRACT FROM AUTHOR]