Improving the Surface Roughness of a Cvd Coated Silicon Carbide Disk by Performing Ductile Regime Single Point Diamond Turning

Silicon carbide (SiC) is one of the advanced engineered ceramics materials designed to operate in extreme environments. One of the main reasons for the choice of this material is due to its excellent electrical, mechanical and optical properties that benefit the semiconductor, MEMS and optoelectronic industry respectively. Manufacture of this material is extremely challenging due to its high hardness, brittle characteristics and poor machinability. Severe fracture can result when trying to machine SiC due to its low fracture toughness. However, from past experience it has been proven that ductile regime machining of silicon carbide is possible. The main goal of the subject research is to improve the surface quality of a chemically vapor deposited (CVD) polycrystalline SiC material to be used in an optics device such as a mirror. Besides improving the surface roughness of the material, the research also emphasized increasing the material removal rate (MRR) and minimizing the diamond tool wear. The surface quality was improved using a Single Point Diamond Turning (SPDT) machining operation from 1158nm to 88nm (Ra) and from 8.49μm to 0.53μm (Rz; peak-to-valley). INTRODUCTION Silicon carbide is used in specialized industries due to its excellent mechanical properties such as extreme hardness, high wear resistance, high thermal conductivity, high electric field breakdown strength and high maximum current density. The fully dense cubic (beta) polycrystalline silicon carbide CVD coating (≈250μm thick) is a potential candidate to be used as mirrors for surveillance, high energy lasers (such as airborne laser), laser radar systems, synchrotron x-ray, vacuum ultraviolet (VUV) telescopes, large astronomical telescopes and weather satellites. The primary reasons CVD coated silicon carbide is preferred for these applications is that the material possesses high purity (>99.9995%), homogeneity, density (99.9% dense), chemical and oxidation resistance, cleanability, polishability and thermal and dimensional stability. Machining silicon carbide is extremely challenging due to its extreme hardness (≈27 GPa) and brittle characteristics. Besides the low fracture toughness of the material, severe tool wear of the single crystal diamond tool also has to be considered. Previous researchers have successfully been able to precisely grind CVD-SiC (using high precision grinding) but this process is very expensive and the fine abrasive wheels often result in an unstable machine/process. Single point diamond turning (SPDT) was chosen as the material removal method as SPDT offers better accuracy, quicker fabrication time and lower cost when compared to grinding and polishing. Although SiC is naturally brittle, micromachining this material is possible if sufficient compressive stress is generated to cause a ductile mode behavior, in which the material is removed by plastic deformation, instead of brittle fracture. This micro-scale phenomenon is also related to the High Pressure Phase Transformation (HPPT) or direct amorphization of the material. 5 The plastic deformation or plastic flow of the material, at the atomic to micro scale, occurs in the form of severely sheared` machining chips caused by highly localized contact pressure. Figure 1 shows a graphical representation of the highly stressed area that result from ductile regime machining. Figure 1: A ductile cutting model showing the high compressive stress and plastically deformed material behavior in brittle materials. A critical depth, dc is determined before any ductile mode machining operation is carried out. Any depth beyond or exceeding the critical depth, also known as the Ductile to Brittle Transition (DBT) depth, will result in a brittle cut. The primary purpose of this research was to improve the surface roughness of a CVD coated SiC disk by SPDT in the ductile regime. The experimental process and results of this investigation will be discussed in this paper. Several parameters such as feed rate, spindle speed and depth of cut were changed to determine the optimum machining conditions to achieve the desired surface roughness, MRR and acceptable tool wear. Processes and procedures to minimize tool wear will be discussed as it is one of the major factors of this machining operation. Before the actual machining was carried out, an experimental test matrix was designed based on previous experiments and some preliminary calculations. Since the equipment used (Micro-Tribometer by CETR) was a load controlled (and not a depth controlled) machine, thrust force calculations were carried out for corresponding required depths of cuts. The Blake and Scattergood ductile regime machining model (as shown in Figure 2) was used to predict the required thrust force for a desired depth of cut. In this model it is assumed that the undesirable fracture damage (which extends below the final cut surface) will originate at the critical chip thickness (tc), and will propagate to a depth, yc. This assumption is consistent with the energy balance theory between the strain energy and surface energy. 7 Figure 2: Model for ductile regime machining. In general, the ductile-to-brittle transition (DBT) is a function of various variables such as tool geometry (rake and clearance angle, nose and cutting edge radius), feed rate, cutting speed and depth of cut. EXPERIMENTAL METHOD The equipment used to carry out all of the machining experiments was the Micro-Tribometer (UMT) from the Center for Tribology Research Inc. (CETR). This equipment was developed to perform comprehensive micro-mechanical tests of coatings and materials at the micro scale. Figure 4 shows the equipment setup for the 6” CVD coated SiC disk (the similar setup was used for the polished 2” SiC disk). A single crystal diamond tool with a 3mm nose radius, -45 degree rake angle and 5 degree clearance angle was used for the cutting tests. The MASTERPOLISH 2 Final Polishing Suspension (contains alumina and colloidal silica with a pH ~9) from Buehler, Inc. was used as the cutting fluid for all experiments involving diamond turning SiC. Figure 3: Machining Setup for the 6” CVD coated SiC disk Final machining of a 6” CVD coated SiC Results from previous experiments which involved scratching and SPDT of CVD SiC, were used as guidelines to establish the experimental plan for an initial set of machining tests. 2,9 Several tests were conducted on polished and as received CVD-SiC to observe the equipment stability, tool condition after machining, thrust force and depth of cut correlation and surface finish of the workpiece after machining. One of the first steps in establishing the machining parameters for the 6” disk was to measure the as received surface roughness (Ra and Rz (peak-to-valley)) using a surface profilometer. Since the target Ra value for the final machined part was below 100nm, it is important to realize that this cannot be achieved in a single pass (as received Ra ≈ 1.158 μm and Rz ≈ 8.486 μm) in order to avoid brittle fracture. This is due to the limitation of the ductile-to-brittle transition depth of the material that cannot be exceeded at anytime, in order to maintain ductile regime machining and avoid brittle fracture that could degrade the surface. The initial goal was to reduce the peak-to-valley (Rz) values of the workpiece. The actual depth of cut is always expected to be less than the programmed depth of cut (in most cases for SiC the actual depth of cut is about half of the programmed depth of cut) due to the elastic properties of the material and tool system. At this depth of cut, the tool would hold up well (the tool did not chip or break), and the cut would still be in the ductile regime and the cuts will not cause additional valleys or cracks to be generated, in addition to what was already in the as received disk. Once reliable data was obtained from the preliminary tests, the machining parameters for the final machining on the 6” SiC disk were determined. Table 1 shows the planned machining parameters for the 6” CVD coated SiC. Table 1: Machining parameters for the 6” CVD coated SiC Pass # Programmed Depth Feed Spindle Speed Fz (Thrust Force) 1 2μm 30μm/rev 6rpm 1287.07mN 2 2μm 30μm/rev 6rpm 1287.07mN 3 2μm 30μm/rev 6rpm 1287.07mN 4 2μm 30μm/rev 6rpm 1287.07mN 5 2μm 30μm/rev 6rpm 1287.07mN 6 2μm 5μm/rev 12rpm 1287.07mN 7 2μm 5μm/rev 12rpm 1287.07mN 8 2μm 5μm/rev 12rpm 1287.07mN 9 500nm 5μm/rev 12rpm 81.32mN 10 250nm 5μm/rev 12rpm 20.99mN 11 1μm 1μm/rev 60rpm 320.79mN 12 500nm 1μm/rev 60rpm 81.32mN 13 500nm 1μm/rev 60rpm 81.32mN 14 500nm 1μm/rev 60rpm 81.32mN 15 500nm 1μm/rev 60rpm 81.32mN 16 250nm 1μm/rev 60rpm 20.99mN A new/relapped diamond tool was used at the start of every pass. The workpiece surface and tools were measured and imaged after every pass to determine the tool condition and to measure the tool wear. For every feed rate used, the spindle rpm was at its slowest possible speed to minimize any tool vibration. Once again the thrust force (Fz) values are an input parameter to obtain the desired depth of cut. RESULTS Final machining of a 6” CVD coated SiC Based on the results from all the test experiments, the final machining was carried out. The results for the final machining experiment are discussed in this section. The overall results are consistent with the preliminary experiments done whereby the maximum improvement in surface finish was after the first pass, i.e. the roughing pass. All spindle speeds were chosen to be the slowest possible for the desired feed. The slowest speeds were chosen in order to eliminate any vibration from the spindle and tool that could make the surface finish worse. Once there is no significant improvement measured at one particular feed, the feed rate is then reduced. The depth of cut for a particular pass was chosen based on the remaining Rz value from the previous pass. The depth of cut was chosen so that it will not exceed the current Rz value and also it must be less than the DBT of the material (≈550nm). The initial eight passes yielded in actual depths of cuts that are larger than the DBT depth of the material. This is a good indication of some brittle mode machining, however, from Figure 2, it is understood that any micro-cracks that do not extend beyond the surface damage depth (yc) for that depth of cut will not make the surface roughness worse. To help understand the effects of certain parameters on the surface roughness of the workpiece several other charts are provided (Figures 4, 5 and 6). Surface Roughness (Ra) vs. Feed