An Atomistic-Information-Based Continuum Inhomogeneous Material Model for Metal Nanorod

H. A. Wu,X. X. Wang
DOI: https://doi.org/10.1007/1-4020-5370-3_351
2006-01-01
Abstract:Mechanical behaviors of materials and structures at nano scale are essentially different from those at macro scale, resulting from surface effect, size effect and time scale effect. To correctly predict the structure-property relations of elemental nanocomponents are very important for the design of nanodevices. Atomistic simulations have been widely used to investigate nanomechanics. The equivalent elastic modulus of a copper nanorod under extension can be obtained using molecular dynamics simulation [1]. In our previous work, it was found that the correct deflection could not be obtained when above equivalent elastic modulus was used to predict the bending behaviors of nanorod [2]. The error was even up to 50%, compared with the direct atomistic simulation. The aim of this work is to investigate the original mechanism of above discrepancy and to present a novel continuum model to predict the bending modulus correctly. We owe this difference mainly to the surface effect. The ratio of surface atoms to the totality is about 60% for copper nanorod with cross-section size of 2nm. In some approximation, the nanorod can be considered as continuum. However, it is not homogeneous across the section. In our continuum model, the nanorod is considered as inhomogeneous material. The material constants of surface atoms and inner atoms are calculated from atomistic simulation, so our material parameters for continuum model of metal nanorod are based on the atomistic information. A finite element analysis of bending is carried out. The result agrees with direct atomistic simulation well, which validates the continuum nanorod model. Further work is to incorporate the research output as a new material model into commercial finite element software. With the development of accurate inter-atomic potentials for a range of materials, classical MD simulations have become prominent as a tool for elucidating the mechanical behaviors of nano-structures. However, the length and time scales that can be probed using MD are still fairly limited. Our continuum metal nanorod model includes nano-effects and supplies another way to study nanomechanics.
What problem does this paper attempt to address?