A simple and efficient methodology is proposed for computing the nonlinear stress-strain behavior and fracture strength under tensile loading of graphene-silicon nanosheet composites. A nonlinear finite-element model (FEM) is developed to obtain the stress-strain relationships, which in general are computed using the computationally expensive procedure of molecular dynamics (MD) simulations. The modelling method, based on the interconnection of continuum micro-mechanics approaches, incorporates bonded contact at the graphene-silicon interface with the graphene being modelled as the multilinear elastic and the silicon assumed as an isotropic material is fed into an atomistic progressive fracture model. Using this model, we get the nonlinear behaviour of graphene, silicon and its composites, as a nonlinear stress-strain curve with their critical stress at inflection point leading to failure. The obtained results of the stress-strain curves, the elastic modulus and the critical stresses of single layer graphene (SLG), silicon nanosheet and their composites (GSNC) with different thickness of silicon nanosheet are in good agreement with the MD results available in the literature. These results elucidate that the tensile strength and Young's modulus of the silicon nanosheet increase enormously by putting the graphene layer on the top and bottom surface of the silicon nanosheet. This model is then employed to study the non-linear stress behaviour for the different orientations of silicon with varying thickness and their corresponding composites with graphene and the effect of various chirality of SLGs are also considered. Thus, we try to establish FEM as a reliable numerical method to obtain the mechanical behaviour of graphene-silicon nanosheet composites which for long-time has been solved using the classical approach of MD. The parametric study which is done using the developed FEM model shows that the GSNC helps to enhance the mechanical properties of silicon nanosheet, which in turn is helpful for the silicon-based semiconductor industry. This fact correlates well with the other methods, well-established in the literature. © 2018 Elsevier Ltd