Efficient and Reliable Nanoindentation Simulation by Dislocation Loop Erasing Method

  • PDF / 4,444,181 Bytes
  • 14 Pages / 595.276 x 790.866 pts Page_size
  • 38 Downloads / 201 Views

DOWNLOAD

REPORT


ISSN 1860-2134

Efficient and Reliable Nanoindentation Simulation by Dislocation Loop Erasing Method Fei Shuang1,2

Pan Xiao1

Yilong Bai1

1

( LNM, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, China) (2 Department of Mechanical and Aerospace Engineering, University of Florida, Gainesville, FL 32611, USA)

Received 2 February 2020; revision received 3 June 2020; Accepted 8 June 2020 c The Author(s) 2020 

ABSTRACT Nanoindentation is a useful technique to measure material properties at microscopic level. However, the intrinsically multiscale nature makes it challenging for large-scale simulations to be carried out. It is shown that in molecular statics simulations of nanoindentation, the separated dislocation loops (SDLs) are trapped in simulation box which detrimentally affects the plastic behavior in the plastic zone (PZ); and the long-distance propagation of SDLs consumes much computational cost yet with little contribution to the variation of tip force. To tackle the problem, the dislocation loop erasing (DLE) method is proposed in the work to alleviate the influence of artificial boundary conditions on the SDL–PZ interaction and improve simulation efficiency. Simulation results indicate that the force–depth curves obtained from simulations with and without DLE are consistent with each other, while the method with DLE yields more reasonable results of microstructural evolution and shows better efficiency. The new method provides an alternative approach for large-scale molecular simulation of nanoindentation with reliable results and higher efficiency and also sheds lights on improving existing multiscale methods.

KEY WORDS Nanoindentation, Dislocation loops, Boundary effects, Molecular statics

1. Introduction Nanoindentation is an effective technique not only to characterize mechanical properties of materials like hardness and elastic response at high spatial resolution in experiments [1], but also to gain insights into plastic mechanisms at atomic level by molecular simulations [2]. In essence, nanoindentation is a typical multiscale problem involving mechanical phenomena at several scales. Firstly, the instability of individual atoms under the indenter tip leads to dislocation nucleation [3]. Afterward, the growing dislocation network constitutes the plastic zone (PZ) which accounts for the hardness, while some dislocations are emitted into the deep of substrate or move along surfaces as separated loops [4]. The expanding PZ reveals the material plasticity, and size effects emerge due to the evolution of dislocation density [5]. Nanoindentation can be directly simulated with empirical-based atomistic methods, such as molecular dynamics (MD) or molecular statics (MS). The computationally intensive nature of nanoindentation, however, limits the simulation scale to a size much smaller than that in experiments. Therefore, developing new methods to simulate nanoindentation with higher reliability and efficiency is of great importance for the research community. Over the past two decades, various simula