13th International Conference on Fracture June 16–21, 2013, Beijing, China -5- Figure 5 (a) and (b) Deformation dependence of the ratio of the transverse and longitudinal strains, 110 110 / ε ε <− > < > − and 001 110 / ε ε < > < > − for the 3D networks under uniaxial tension, respectively. (c) and (d) Ratio of transverse and longitudinal strains as a function of CNT segment length N at 110 ε< > = 0.01. Visualizing the detailed structural evolutions of this 3D nano-truss provides insights on the structural and mechanical properties, such as the deformation mechanism, the negative Poisson’s ratio and fracture modes. Figure 6 displays straining direction dependence of the representative atomic structural development of the 3D nano-truss. For N = 0, the initial side-view shows the porosity of the 3D nanostructures, analogous to the zeolite structure [25], offering a possibility for molecular storage and filters. During the deformation, three stages can be identified: firstly, there is a curvature flattening process with the increase of the strain, as presented in the zoomed-in local structures of Figure 6a and 6c, which results in a negative Poisson’s ratio. In this stage, the curvature-flattening dominates the deformation with no irreversible structural changes, triggering an expansion of pore between fullerenes as well as the fullerene cages. Secondly, the subsequent graphitic structure with negative curvature ultimately causes a contraction when the strain approaches a critical value. This is confirmed by the change of Poisson’s ratio from negative to positive as shown in Figure 3. Finally, an additional strain increment leads to a fracture at the weak octagon rings at the junction. The fracture propagates in the plane perpendicular to the strain direction towards complete rupture, in which a number of monoatomic carbon chains formed, similar to what is widely observed in the stretching of pure CNTs. A longer CNT for spanning the fullerenes is exploited to tune the properties of this 3D networks. Particularly, sequential deformation configurations of the 3D network with N = 10 are displayed in Figure 6b and 6d. It can be noticed that the long length of CNTs brings a sparse 3D network, and orientation dependence of
RkJQdWJsaXNoZXIy MjM0NDE=