Numerical simulations and cutting experiments on single point diamond machining of semiconductors and ceramics

This chapter presents numerical simulation work and single-point nano-machining experiments conducted on semiconductor and ceramic materials, e.g. silicon (Si) and silicon carbide (SiC). The apparent ductile mode material removal mechanism observed in these materials is believed to be the result of a high pressure phase transformation (HPPT), which generates a small Correspondence/Reprint request: Dr. John A. Patten, Department of Manufacturing Engineering, Western Michigan University, Kalamazoo, Michigan, USA. E-mail: [email protected] John A. Patten et al. 2 plastic zone of material at the tool-work piece contact interface that behaves in a metallic manner. This metallic behavior provides the motivation for using AdvantEdge, a metal machining simulation software, for comparison to experimental results. The first section of this chapter describes numerical simulations of single point diamond turning (SPDT) experiments conducted on single crystal 6-H SiC. The SPDT experiments have shown ductile chip formation similar to that seen in the machining of metals. The cutting and thrust forces generated from the experiments under ductile cutting conditions compare favorably with the machining simulation results. As the depth of cut is decreased (from 250 nm to 100 nm down to 50 nm) the material removal mechanism transitions from brittle to a ductile response, with the 50 nm cuts being dominated by ductile events. Thus the forces from the experiment and the simulations are in much better agreement for the smaller depths of cut, i.e. below the critical depth of cut that establishes the ductile to brittle transition (DBT), as ductile conditions exist in both the simulation and experiments. The differences in the results are due to elastic deformation under the tool leading to increased rubbing at small depths of cut, and the occurrence of brittle behavior for the experimental results at the largest depths of cut. The first section of this chapter also presents a predictive analytical model for machining with a large negative rake angle tool, the condition which is very common in nano machining. The second section presents results from a fly-cutting experiment conducted on single crystal 6H-SiC. The cutting tests were performed to determine the critical depth of cut (for ductile to brittle transition), while machining on the (0001) face. The cuts were made using a single crystal synthetic diamond tool with a -45° rake angle and a 1 mm nose radius at a speed of 82.5 mm/sec. A scanning white light interferometric microscope (Wyko RST) is used to detect the ductile to brittle transition of SiC based on the surface topography of the cut surfaces. For each individual cut, the cutting and thrust forces are correlated to the resulting depths along the cut, to determine the chip cross-sectional areas. This information is then used to normalize the force values, which effectively eliminates variations in force due to problems with workpiece flatness while machining. An estimate of the friction factor due to interaction of the SiC and diamond tool is also obtained. The third section presents 3-D numerical simulation work relating to scratching experiments conducted on both Si and SiC with a stylus-radius diamond tip. These experiments were also aimed at determining the DBT depth in these materials. The simulations were conducted at the different depths and the cutting and thrust forces were compared to experimental Simulations and machining experiments on ceramics 3 results. The initial attempt at simulation of scratching in 3-D shows promising results. Introduction The mechanical and thermal properties of silicon carbide (SiC) have traditionally allowed for its use in refractory linings and heating elements for industrial furnaces, as an abrasive in manufacturing processes, and in wear resistant parts in rotating machinery such as pumps and engines. There is currently growing interest in the use of SiC in the optics industry for space based laser mirrors, as a replacement for beryllium. Also in the electronics industry SiC is being used for high-powered/high temperature devices, where its high thermal conductivity, high electric field breakdown strength, and high maximum current density make it more promising than silicon (Si) [1]. The successful acceptance of SiC in these industries will require satisfaction of stringent requirements on form accuracy and sub-surface damage [2] where surface finishes better than 10 nm are considered standard specifications [3]. Brittle mode grinding and chemical-mechanical polishing (CMP) have been used to meet product requirements and manufacturing productivity. Since grinding and polishing are slow and costly there is a continuing motivation to use SPDT to eliminate some of these steps [2] to reduce production time and to reduce overall component costs. Additionally, traditional grinding induces micro-cracks in the surface of brittle materials [4], while polishing is required to then remove the surface and subsurface damage caused by grinding [2]. Ductile mode machining or cutting technology has been studied as a replacement for grinding of optical devices and semiconductors, such as Si [2, 4]. The machining of this hard and brittle material is presumed to be made possible due to the ductility, and presumed metallic nature, of its high pressure phase (HPP), or possibly as a result of amorphization [15]. The tremendous hydrostatic pressure and shear stress in the region immediately surrounding the tool cutting edge allows materials like Si and also SiC to behave in a ductile fashion at small depths of cut, exhibiting plastic deformation at room temperatures [6]. The ability to characterize the mode of material removal through simulations provides an alternative approach to understanding the effect of varying machining parameters, such as rake angle, cutting edge radius, and depth of cut. A comparison of the resultant force per unit cross-sectional chip area (expressed as a pressure in GPa) can provide a measure or an indication of the mode of material removal, i.e. ductile or brittle. John A. Patten et al. 4 1. Comparison between numerical simulations, cutting experiments and an analytical model for single point diamond turning of single crystal silicon carbide 1.1 Background Experiments conducted in recent years on Si [2, 4] and silicon nitride [5] have demonstrated a ductile mode of material removal similar to that seen in metals. Ductile mode implies a process dominated by ductile or plastic material deformation and removal rather than brittle fracture, resulting in a smooth surface (similar to a polished surface) free of fracture damage [7]. The ductile nature of these hard and brittle materials has been attributed to the HPPT, which occur at or near room temperature. The HPPT is created as a result of the contact between the sharp tool and the workpiece at or below the critical depth of cut [6]. This critical depth has been demonstrated to be in the nanometer range for SiC [2, 5, 6, 7]. Recent work [7] involving Single Point Diamond Turning (SPDT) of single crystal SiC was conducted to study, among other things, the effect of varying the rake angle and the depth of cut on the cutting and thrust forces, the resultant surface finish and the ductile to brittle transition (DBT). The forces from the experiment directly indicate whether the material removal is ductile or brittle during processing, while the final surface finish roughness indicates the achievement of ductile or brittle material removal. Ductile machining is characterized by higher cutting forces, as it takes more energy to remove material in a ductile mode as opposed to brittle material removal for the same volume of material [7]. The focus of this section is to determine and evaluate the capability and accuracy of predicting the experimental results in the work reported in [7] based on 2-D turning simulations conducted using version 4.5 of the commercial machining simulation software AdvantEdge. 1.2 Experimental procedures and conditions Results from three sets of experiments on single crystal SiC are considered for comparison to the simulations. The first set of experiments involved depths of cut (in-feed) of 100 nm, 300 nm and 500 nm using a 2 mm round nose mono-crystalline diamond tool with a 0° rake angle. This same tool was reoriented to create an effective rake angle of -45°, and cuts were made at the same depths forming the second set of experiments. The first two sets of experimental results are reported in [7]. Additionally, machining at a depth of 50 nm was conducted but not reported previously. The third set of experiments involved 50 nm and 250 nm depths using a flat nose mono-crystalline diamond tool with a -45o rake angle. The experiments were all conducted at a low speed Simulations and machining experiments on ceramics 5 of 3 m/min to minimize temperature rise and consequent thermal effects, such as thermal softening of the material. Dry cutting conditions allowed for collection of machining chips and provided simplified simulation boundary conditions, i.e. no coolant or cutting fluid effects were considered. Table 1 and Table 2 summarize the experimental process parameters. Table 1. Experimental Set I and II. Feed (nm) Spindle speed (RPM) Speed (m/min) Tool geometry (0° rake)/round