The evolution of machining-induced surface of single-crystal FCC copper via nanoindentation
© Zhang et al.; licensee Springer. 2013
Received: 26 February 2013
Accepted: 7 April 2013
Published: 4 May 2013
Skip to main content
© Zhang et al.; licensee Springer. 2013
Received: 26 February 2013
Accepted: 7 April 2013
Published: 4 May 2013
The physical properties of the machining-induced new surface depend on the performance of the initial defect surface and deformed layer in the subsurface of the bulk material. In this paper, three-dimensional molecular dynamics simulations of nanoindentation are preformed on the single-point diamond turning surface of single-crystal copper comparing with that of pristine single-crystal face-centered cubic copper. The simulation results indicate that the nucleation of dislocations in the nanoindentation test on the machining-induced surface and pristine single-crystal copper is different. The dislocation embryos are gradually developed from the sites of homogeneous random nucleation around the indenter in the pristine single-crystal specimen, while the dislocation embryos derived from the vacancy-related defects are distributed in the damage layer of the subsurface beneath the machining-induced surface. The results show that the hardness of the machining-induced surface is softer than that of pristine single-crystal copper. Then, the nanocutting simulations are performed along different crystal orientations on the same crystal surface. It is shown that the crystal orientation directly influences the dislocation formation and distribution of the machining-induced surface. The crystal orientation of nanocutting is further verified to affect both residual defect generations and their propagation directions which are important in assessing the change of mechanical properties, such as hardness and Young's modulus, after nanocutting process.
Ultraprecision machining at nanometric scale is increasingly required in micromachining and nanomachining to produce parts of intricate features and surface finish quality. Material removal at such a small uncut chip thickness involves subsurface deformation, and in conventional cutting, the effect of subsurface deformation is neglected as the uncut chip thickness is significant. However, it is not the same case in nanocutting due to the small uncut chip thickness on the order of several nanometers or less. Thus, the effect of subsurface deformation should not be neglected as the uncut chip thickness is in the same scale. Subsurface deformed layer is related to the deformation and damage in the material especially in the micro- and nanoscales, in which not only the size is reduced substantially but also the physical characteristics on optics and electricity of the material become different. Recently, the mechanisms of subsurface deformation have become the key issues to be investigated.
Many investigations have been conducted to study the subsurface deformed layers during nanocutting process via molecular dynamics (MD) simulations. Shimada and Ikawa et al. performed MD simulation of microcutting of free machining materials under perfect motion of a machine tool. Based on the radial distribution function, they found that the ultimate depth of the deformed layer of a specimen is 5.0 nm[3–5]. Zhang et al. conducted MD simulation of nanometric cutting on single-crystal copper. A new criterion based on single-atom potential energy variation was established. However, the previous studies evaluated the subsurface atom deformation behaviors mainly by studying and analyzing the cutting forces and potential energy variations. Although these features of different deformation behaviors can be revealed efficiently, the potential energy variation of atoms is hardly measured by current experimental equipment. Therefore, it is an important issue to investigate the surface properties of the subsurface deformed layers after nanocutting process.
Nanoindentation is the most frequently used technique to measure surface properties such as Young's modulus and hardness. Investigations on exploring the performances of friction and wear of single-crystal materials are thus of scientific and technological interest. For this reason, a lot of studies on nanoindentation based on experimental and various theoretical models have been carried out to have deep understanding of the performance of these surface and near-surface tribological properties. Yan et al. performed nanoindentation tests on ultraprecision diamond-turned silicon wafers[7, 8]. Compared with those of pristine silicon wafers, the machining-induced amorphous layer was with significantly higher microplasticity and lower hardness than pristine silicon. Zhao et al. performed the same process and analyzed the machinability of the material and its structure via molecular dynamics simulation. Although the experimental and theoretical results revealed the structure transformation in diamond semiconductors, the mechanism of the phase transformation did not suit for most of metal materials. Since the lattice structure of a metal is different from a semiconductor, the phase transformation is not fitful for most face-centered cubic (FCC) metals. Consequently, understanding of the different performances and machinability of the machining-induced layer in a FCC metal becomes essential.
In this paper, theoretical analysis and investigation on the properties of subsurface deformed layers in nanocutting process with the aid of nanoindentation test will provide much information on the mechanisms of the deformation in the material. The displacements of dislocations are simulated to have better understanding of the mechanism of the damaged layer in nanocutting and nanoindentation test on a machining-induced surface. The remainder of this paper is organized as follows: The ‘Methods’ section gives the models and conditions of the MD simulation. The ‘Results’ section presents the results of the simulation and discusses the results in detail. The ‘Discussion’ section discusses the effect of cutting directions along different crystal orientations on the subsurface deformed layers. The last part draws some interesting conclusions.
The diamond tool consists of 21,823 carbon atoms, and the rake angle and clearance angle are 0° and 7°, respectively. For conventional cutting processes, the cutting tool edge is usually assumed to be very sharp, that is, the edge radius is negligible. However, during nanocutting process of materials, this assumption is not reasonable since the cutting tool edge radius is on the same scale as the undeformed chip thickness. Thus, the simulation has been done with the cutting edge radius of 2 nm. The spherical indenter contained 36,259 atoms with a radius of 50.0 Å.
where a ix represents the i atom's acceleration in the X direction, m i is the mass of the i atom, F ix is the interaction force between the i atom by the j atom in the X direction, x i indicates the i atom's X-coordinate, and V is the potential energy.
where N is the number of atoms in groups, v i represents the velocity of the i atom, k b is the Boltzmann constant which is equal to 1.3806503 × 10−23 J/K, and T represents the temperature on atoms.
In this paper, there are two kinds of atoms in the MD simulation model, which are C and Cu atoms. Therefore, there are three different atomic interactions between them, which are the interaction between single-crystal copper atoms (Cu-Cu), the interaction between diamond atoms (C-C), and the interaction between copper atoms and diamond atoms (Cu-C) or (C-Cu).
where ρ i (r ij ) is the contribution to the electronic density at the site of the atom i, and r ij is the distance between the atoms i and j.
Because diamond is much harder than copper, the diamond tool and indenter are both treated as a rigid body in the simulation. Therefore, the atoms in the tool are fixed to each other relatively, and no potential is needed to describe the interaction between diamond atoms (C-C).
Parameters in the standard Morse potential
r 0 (Å)
In order to reduce the boundary effect and size effect, the model scale should be large. As a result, the simulation becomes computationally expensive. To avoid these problems, the periodic boundary condition is set along the Z direction. The specimen surface of the X-Z plan is machined, so it is a free surface. Both the diamond tool and the diamond indenter are set as a rigid body. This was followed by an energy minimization to avoid overlaps in the positions of the atoms. The simulation model was equilibrated to 296 K under the microcanonical (NVE) ensemble, and the initial velocities of the atoms were assigned in accordance with the Maxwell-Boltzmann distribution.
Computational parameters used in the MD simulation model
Tool: diamond (rigid)
Indenter: diamond (rigid)
EAM potential function
75a × 35a × 50a
Rake angle, 0°
(a is the lattice constant, 0.3614 nm)
Clearance angle, 7°
Radius, 50.0 Å
Number of atoms
[ī00] on (010) surface
 on (010) surface
The three-dimensional MD simulations were performed by the large-scale atomic/molecular massively parallel simulator (LAMMPS)a developed by Plimpton et al.[11, 15]. The parallel computation was realized under the help of message passing interface library.
According to Figure 3a, there are several different defects generated during the nanocutting process. Various defects distributed in the specimen are marked by the numbers in Figure 3a. The single vacancy, marked with number 1, is easily identified by its simple dependent structure and atomic coordinated number. When the dependent single vacancies are gathered by the movements and interactions of dislocations in the specimen, the immobile vacancy clusters and a vacancy chain (marked with 2 and 3 in Figure 3a) are generated beneath the machining-induced surface, which may largely alter the mechanical properties of the machined surface after nanocutting.
The threading dislocation, marked with number 4, belongs to one of the mobile defects in the specimen. It is well shown that the threading dislocation marked in the specimen is parallel with the slip vectors associated with the FCC (111) surface. According to the position-sensitive criterion, its motion in the specimen under the machining-induced surface determines the plastic deformation of the material in nanocutting.
The dislocation loop of numbers 5 and 6, which was emitted from the tool-specimen interface, denotes the dislocation loops. Unlike the single vacancy defects distributed in the specimen, the dislocation loops glide along with the movement of the diamond tool. In addition, the motion directions of the dislocation loops are not the same. Some dislocations penetrate into the specimen towards the bottom surface, while others are moving along the cutting direction beneath the machining surface. Their motivation promotes not only the nucleation of other defects in the specimen but also theirselves. They initially generated from one side of the specimen and finally went inside the opposite site of the boundary.
Figure 3c provides some different views of the new generated surface. Some dislocation can be seen on the surface. It is also seen that the dislocations on the machining surface marked with numbers 7 and 8 are parallel with the slip vectors [ī0ī] and [ī01]. The two directions in the specimen are the easiest glide vectors in the surface. Many generated dislocations are involved in the accumulated atom pile-up in front of the diamond tool. The black arrow in the figures indicates the cutting direction. Some defects remained on the machining-induced surface, marked with numbers 9 and 11 in Figure 3c. The vacancy-related defects on the machining-induced surface, number 9, are not only immobile but are also located limited on the surface, while the dislocation-related defects are completely contrary. The dislocation loop is usually distributed along with such a defect on the surface. The dislocation nucleation and escape in submicrometer single-crystal FCC metal materials have been observed and proven in some previous studies using experiments and simulations[18, 19].
Figure 4 (a1 and b1) shows the atomic total energy distribution and kinetic energy distribution of the initial surface, and Figure 4 (a2 and b2) shows those of the relaxed machining-induced surface. According to Figure 4, there is no obvious difference in energy distribution on both the relaxed machining-induced surface and the initial surface. Although more high-energy defects are observed to be distributed on the relaxed machining-induced surface (marked with black circles), the overall surface condition is about the same with the initial surface. The result implies that the relax stage after the nanocutting process is well performed for the atomic total energy distribution and that kinetic energy on the surface returns to a low and stable situation. Since the atomic total energy and kinetic energy are about the same as those of the former initial surface, the influential factors due to different energy distributions are well excluded.
The evolution of interior defects inside the specimen during nanoindentation governs the mechanical properties of the surface, especially the hardness and Young's modulus. Therefore, the investigation of the nucleation and penetration of dislocations beneath the indenter seems strongly necessary.
Figure 5 (a1 to a4, and b1 to b4) shows a typical material deformation process of pristine single-crystal copper during nanoindentation with a series of structure evolutions for nucleation of initial defect beneath the indenter. In the undeformed state, none of defects are distributed or generated beneath the indenter. With small deformation, a few vacancies generate just beneath the indenter, which marks the beginning of nucleation of dislocations. As the single-crystal copper atoms experience the displacive structure transition, the well-known dislocation embryos are gradually developed from the sites of homogeneous nucleation as shown in the prospective close-up view of Figure 5 (b4). In addition, the atomic glides on the surface are also clearly marked with black arrows, which are parallel with the slip vectors associated with the FCC (111) surface. The motivation of these glides indicates the displacive plastic deformation around the indenter as shown in Figure 5 (b4).
Showing contacts to the nucleation of dislocations in the pristine single-crystal copper, the process in the subsurface of the machining-induced surface is different. Figure 5 (c1 to c4, and d1 to d4) presents a universal process of the dislocation evolution in the subsurface with initial imperfection of the machining-induced surface. Before the indenter penetrates into the machining-induced surface, there have been some vacancy-related defects distributed on the surface as shown in Figure 5 (c1 and d1). When the indenter penetrates into the surface, the dislocation embryos are immediately developed from the vacancies around the indenter. Although the glide directions of such defects are still along slip vectors associated with the FCC (111) surface, the initial vacancy-related defects distributed on the machining-induced surface become the beginnings of mobile dislocation loops. The formation energy of mobile dislocation of such a process is largely reduced. In addition, much more dislocation loops in the specimen are motivated by the indenter-specimen interaction, leading to the permanent plastic deformation of the material.
In Figure 7, the loading curves of the two surfaces present some different characteristics. The discontinuity can be clearly observed as for the copper with perfect structure, which agrees with conventional studies. However, the loading curve of the machining-induced surface is much smooth. The differences are due to the dislocation nucleation-induced elastic and plastic deformation transformation. Compared to the maximum energy needed to be developed and propagated in the machining-induced surface, it is much larger in the pristine copper specimen. Since the high-energy initial defects have existed on the machining-induced surface, the power to trigger dislocation nucleation is less needed. When the dislocations emit from the dislocation nucleation and propagate in the specimen, the accumulated energy is released. Therefore, the amplitude value of the indentation curve on the pristine surface is much larger than that on the machining-induced surface.
where ϵ is a constant and depends on the geometry of the indenter (ϵ = 0.72 for cone indenter, ϵ = 0.75 for paraboloid of revolution, and ϵ = 1.00 for flat indenter). hmax is the maximum penetration depth, and S is the contact stiffness. Ac is the projected contact area under the peak indentation depth.
where β is a constant and depends on the geometry of the indenter (β = 1.034 for a Berkovich indenter, β = 1.012 for a Vickers indenter, and β = 1.000 for a cylinder indenter).
where E and ν are the elastic modulus and Poisson's ratio for the sample; E i and ν i are the elastic modulus and Poisson's ratio for the indenter, respectively. For the diamond indenter, E i = 1,141 GPa and νi = 0.07. The indenter was assumed to be rigid as mentioned above, and the value of E i is infinite; v s is equal to 0.278.
The applied load versus penetration depth in loading stage
Applied load to the indenter (nN)
Table 3 shows the comparison of indentation loads at different penetration depths of the pristine single-crystal copper specimen and machining-induced surface. It can be noted that the indentation loads on the machining-induced surface are much smaller than those on the pristine surface with the same indentation depth, respectively. No remarkable difference was found when the maximum indentation penetration depth is larger than 2.0 nm. The amplitude value of the indentation curve on the pristine surface is much larger than the other. It is due to the dislocation embryos which developed and propagated in the specimen under the diamond indenter. However, when the maximum penetration is smaller than 2.0 nm, the hardness of the diamond-turned surface becomes distinctly lower than that of the pristine copper. At a sufficiently small load, the indentation response will be mainly due to the surface effects. At a slightly larger indentation penetration depth, the indentation loads are much smaller than those of the pristine single-crystal copper surface. It can be concluded from these results that the machining-induced surface is softer than pristine single-crystal copper.
In conventional metal machining, the near-surface layer is much harder than the original material in the surface. Such a surface-hardening phenomenon is due to work-hardening effects. The work-hardening effects are often due to the dislocation winding near the crystal boundary, where the dislocations are not able to pass across the potential barrier. However, the surface-softening effect during machining is due to no crystal boundaries in single-crystal copper, and the dislocation activities are free to move.
It can also be noted that the calculated hardness of the pristine single-crystal copper specimen and machining-induced surface is 10.55 and 9.25 GPa by Equations 5, 6, 7, 8, 9, respectively, and the elastic modulus is 120.4 and 117.7 GPa, respectively. The machining-induced surface has a lower hardness than pristine single-crystal copper by about −12.3%, and the elastic modulus has no significant disparity (about 2.21%). The immobile dislocations on the machining-induced surface serve as the origin of mobile dislocations in the nanoindentation. The permanent plastic deformation is derived from the movement of dislocations. It has been revealed that the machining-induced surface would influence the physical properties of pristine single-crystal copper as well as other single-crystal FCC metals. The dislocations during nanocutting have been shown to play an important role in the formation of interior defects as well as surface profiles. Therefore, the accurate prediction of the thickness and mechanical properties of the machining-induced surface becomes vital when trying to use it in the application.
Previous studies have introduced the concept of the subsurface damage layer after nanomachining. The criterion of the material damage nanocutting has a lot of statements, such as the thickness of the damage subsurface and the variation of potential energy. In fact, the dislocations distributed in the specimen alter the machining-induced surface mechanical properties. The immobile vacancy-related dislocations may lead to the nucleation of mobile dislocations.
The applied load versus penetration depth in loading stage
Applied load to the indenter (nN)
Cutting direction [ī00]
Cutting direction [ī01]
The variations of hardness and Young's modulus of the machining-induced surface with various cutting directions along different crystal orientations are calculated. The hardness of the machining-induced surface along [ī00] and [ī01] is 9.25 and 11.16 GPa by Equations 5, 6, 7, 8, 9, respectively, and the elastic modulus is 117.7 and 126.46 GPa, respectively. The machining-induced surface along [ī00] has lower hardness than the machining surface cutting along [ī01] by about −17.1%, and the elastic modulus has no significant disparity (about 6.9%). The comparison demonstrates that they are in excellent agreement with the anticipation that the cutting force along the different cutting directions on the same surface is not the same. Larger cutting force causes more severe damage in the subsurface, leading to more changes of the properties of the machined surface.
The present investigation has shown how the machining-induced surface affects the mechanical properties in the atomic level of single-crystal copper by molecular dynamics simulation. Based on the above analysis, some interesting conclusions can be drawn as follows.
Hybrid potentials including the Morse and EAM potentials were employed to simulate the nanoindentation test on the machining-induced copper surface. The nanocutting simulation was carried out at the nanocutting velocity of 200 m/s. The simulation results show that some kinds of defects remain in the subsurface of the machining-induced surface. The defects in the damaged layer alter the mechanical properties of the machining-induced surface. When the indenter penetrated into the machining-induced surface after an adequate relaxation, the dislocation embryos derived from the vacancy-related defects are distributed in the subsurface. These results show that the hardness of the machined surface is smaller than that of single-crystal copper. In addition, the hardness and Young's modulus are calculated from the simulation results, which further verify the former analysis according to the motivation of dislocations in the specimen.
Then, the nanocutting was performed along different crystal orientations on the same crystal surface. It is shown that the crystal orientation directly influences the dislocation formation and distribution in the machining-induced surface. The crystal orientation of nanocutting is further verified to affect both dislocations and residual defect generations that are important in assessing the change of mechanical properties after nanocutting in this length scale.
aDistributed by Sandia National Laboratories, Albuquerque, NM, USA.
This research is funded by the National Natural Science Foundation of China (grant no. 50905073, 51275198, 51105163), Special Projects for Development of National Major Scientific Instruments and Equipments (grant no. 2012YQ030075), National Hi-tech Research and Development Program of China (863 Program) (grant no. 2012AA041206), Key Projects of Science and Technology Development Plan of Jilin Province (grant no. 20110307), and Graduate Innovation Fund of Jilin University (grant no.20121084).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.