Nanometric cutting mechanism of silicon carbide

Abstract

Ductile to brittle transition is critical to achieve nanometric surfaces in the ultraprecision diamond cutting of silicon carbide. Although atomic simulations have long been used to better understand this mechanism, the extremely small model scale limits its capability in matching the actual cutting process. To overcome this serious issue, an enhanced molecular dynamics method is proposed in this study, which successfully predicts and clarifies the onset of brittle regime machining, and indicates the essential roles of dislocation and the shear band. The experimental results validate the effectiveness of this modelling approach.

Introduction

The widespread use of silicon carbide (SiC) in advanced electronic devices and optical systems has led to a large demand for high surface integrity with nanometric finishing. Apart from grinding and polishing, diamond cutting is considered to be a more attractive approach due to its higher efficiency and ability to obtain complex geometries. However, the great hardness and brittleness of this material are critical issues that influence the surface quality and tool life, which necessitates a good understanding of the cutting mechanism. For SiC, high pressure phase transformation (HPPT) and dislocation activity are two sources of plastic deformation, and tool wear is caused by the high temperature and graphitization induced by abrasive action [1]. In addition, 4H-SiC exhibits lower subsurface deformation and 6H-SiC lower cutting resistance [2]. To improve the machinability, ion implantation can be employed to soften the surface layer [3].

Ductile to brittle transition (DBT) has both theoretical and practical significance in the nanometric cutting of SiC, but is not yet fully understood. Experimentally, the critical undeformed chip thickness (UCT) is the parameter most commonly used to indicate change in the cutting status, and numerical method, especially molecular dynamics (MD), is also necessary to investigate this process at atomic and close-to-atomic scale (ACS) [4]. A key problem in MD research is the small model scale, such as that of the UCT and tool edge radius, compared with the actual cutting process. One solution to this problem is to speed up the computation via hardware such as the graphics processor unit [5]. Another solution is to develop a multiscale method, in which the resolution of the region with large deformation is refined to atomic scale and the remaining regions are approximated as a continuum or as virtual particles at the meso or micro scale [6]. However, use of a constitutive model and information transportation through the hand shaking region would reduce the accuracy, and the random behavior of material during cutting also makes the self-adapt particle refinement difficult. Therefore, the totally atomic description continues to have merit for handling mechanical problems at ACS.

In this paper, we propose an enhanced MD approach specifically designed for the study of nanometric cutting. In this method, the whole workpiece model is replaced by a part of the material that dynamically follows the tool motion, so the total number of atoms is significantly reduced. The continuity of both the material lattice and stress field is well achieved using this newly developed approach. To make a direct comparison and determine the extent to which the capability of MD can be enhanced, simulations at a scale similar to that of experiment are performed. The DBT mechanism of 6H-SiC is then investigated with experimental validation.