Predicting the strength of carbon nanotube reinforced polymers using stochastic bottom-up modeling

Abstract

It is intended to predict the strength of carbon nanotube reinforced polymer. For this purpose, a sequential multi-scale modeling technique is developed taking into account effective parameters at various scales of micro, meso and macro. CNT-polymer interaction, CNT orientation and CNT waviness are considered at the scale of micro, while formation of CNT aggregates is captured at meso-scale. The occurrence of failure in the form of either matrix failure or CNT and polymer debonding is considered in the modeling as the two prominent failure mechanisms under tensile loading. At the uppermost scale of macro, the investigated material region is partitioned into constitutive blocks using regular tessellation technique accounting for inhomogeneous material behavior. The induced uncertainties during the processing of carbon nanotube reinforced polymers necessitate stochastic modeling. The experienced inconsistencies consisting of CNT length, orientation, waviness pattern and intensity of agglomeration are captured using random parameters. Therefore, the constitutive behavior of each block at macro-scale is randomly picked from the outputs of the conducted modeling at lower scales and full stochastic modeling is implemented. The outputs of the developed modeling are compared with available experimental observations in published data and remarkable agreement is observed.