| Περίληψη: | In the present dissertation the effective thermal (keff), the effective elastic modulus (Eeff, Geff) and the effective strength (σmax, τmax) of Carbon Nanotubes (CNTs) Reinforced Polymers (CNTsRPs) were predicted by developing finite element continuum homogenization models. For this purpose innovative representative volume elements (RVEs) that take into account determinant nanoscale factors for the “nanostructure–effective property” relationship in each case were developed. Both the influence of CNTs clustering and CNTs-polymer matrix interphase on the Eeff and Geff were investigated by developing “clustered” and “hybrid” RVEs. Additionally “perfect” RVEs were considered for comparison reasons. The validation of the method was confirmed by the correlation of the predicted values with corresponding experimental. The results reinforced the argument of an interphase formation due to the CNTs functionalization. The influence of the Kapitza resistance RKap on the keff of CNTsRPs was investigated. For this purpose RVEs that take into account both the value of the RKap and its extent were developed. The correlation of the predicted values with corresponding experimental revealed a unique phenomenological Kapitza resistance RKapPh for each one of the CNTs content. The plotting of the RKapPh versus the CNTs content showed a linear increase. Tthis observation was related directly to the increased CNTs “clustering” intensity at higher CNTs contents. The σmax and τmax of CNTsRPs were predicted by considering the “perfect” RVEs developed earlier. To this end a progressive damage material model was developed. The results showed that the CNTs cause significant and similar increase of both Eeff and Geff. While, the τmax increased more than the σmax. The above developed models were implemented for the prediction of the equivalent material model (EMM) of a CNTs reinforced epoxy adhesive (CNTsRAD) that was used for the bonding of a single lap joint (SLJ). The macroscopic “cohesive” failure of the SLJ was modeled by applying the Cohesive Zone Model and following the “local” damage mechanics approach. Reliable experimental data confirmed the validation of the model. Eventually the macroscopic response of SLJ bonded with CNTsRAD was predicted by developing a two step multi-scale modeling approach. The results showed the determinant contribution of the “mechanical” parameters against the “fracture” parameters of the EMM to the macroscopic response of the SLJ.
|
|---|
