Single Point Diamond Turning of CVD coated Silicon Carbide

Scratching experiments, using diamond styli and single point diamond tools, were performed to simulate Single Point Diamond Turning (SPDT). The results of these experiments were used to determine if a ductile response is possible, and then to determine the critical depth of cut or penetration depth for the ductile to brittle transition (DBT). The depths of the scratches produced at different loads were measured and correlated to the ductile and brittle response of the material. The DBT depth for Chemically Vapor Deposited (CVD) coated Silicon Carbide (SiC) samples was determined. The analysis for the critical depth (DBT) did confirm the possibility for SPDT of CVD coated SiC in the ductile regime. These results were further used for SPDT of CVD SiC. Post experimental analysis of the machined surface did reveal a final surface roughness of 8-20nm, successfully demonstrating ductile regime machining of CVD coated SiC. INTRODUCTION Machining of brittle materials has reached a level where the focus is turning towards commercial applications. CVD SiC is an excellent reflective optics material, exhibiting superior polishability with low scatter. Since optics applications are a high priority for this material, the surface roughness of the finished CVD SiC should be less than 60 nm. This level of surface roughness is achievable by processes like polishing and lapping. A problem that persists with polishing or lapping is the resultant figure error, primarily flatness or form error deviations. The larger the diameter or size of the piece, the more difficult it is to maintain the flatness with traditional processes. Hence to overcome the limitations of these processes (polishing and lapping) investigation of single point diamond turning (SPDT) was pursued and is reported in this paper. Determination of the DBT depth of CVD coated SiC (from Poco Graphite) using scratching tests (simulated machining), preceded the actual machining experiments. The detailed explanations of scratching experiments have been divided into different sections corresponding to the experimental plan and materials. Two major sets of experiments were performed to determine the DBT depth using an instrumented micro tribometer. One experiment involved scratching using a diamond stylus and the second one used an inclined plane experiment in which a flat nose diamond tool was used to make a scratch. The inclined plane experiment was conducted only on the Poco Graphite sample. A comparison was made between the resultant surface (ductile/brittle) and the force data derived from the tribometer. The speeds of the scratches were kept slow at 0.005 mm/sec to minimize thermal effects. Our primary goal was to demonstrate that SPDT of brittle materials like SiC is possible. Hence a 6 inch diameter SiC piece was machined and reported on herein. The piece was supplied by Poco Graphite Inc. These experiments were executed using a Universal Multi System Tribometer (UMT) from CETR Inc. Unlike precision lathes, a tribometer is typically used for experiments concerning tribological applications. One difference between precision lathes and the tribometer is that precision lathes are displacement based machines and tribometers are load based devices. Chemical mechanical polishing (CMP) fluid was used as a cutting fluid for these experiments to reduce tool wear. The single crystal diamond tools used for SPDT of CVD coated SiC did not show appreciable wear on their cutting edges. The use of the CMP slurry as a cutting fluid most likely provided this positive benefit, i.e. the fluid acted as a lubricant and possibly provided some chemical enhancement. Future work will involve higher cutting speeds to investigate productivity capabilities and enhancements. The post experimental analysis, which primarily included depth measurements of the scratch, used a white light interference microscope (Wyko RST). The data shown for x and y profiles within the Wyko image were averaged for completely ductile scratch profiles. The DBT and the brittle fracture profiles are not averaged as they have the presence of both ductile and brittle behavior. PRE-SCRATCHING PROCESS The CVD coated SiC as received from Poco Graphite Inc) had a very high surface roughness as shown in Table 1. Initial calculations for DBT depths for CVD SiC revealed the 1 Copyright © 2006 by ASME critical depth to be approximately 40 nm. Hence scratching a rough surface to find a DBT depth of 40-50nm was difficult if not impossible for the as received sample (with surface roughness > 500 nm, i.e. more that 10x the estimated DBT depth). Secondly, imaging was a problem with the high surface roughness, as it is difficult to image dull surfaces under white-light microscopes, which we were using for determining the scratch depths and confirm/determine the DBT. Thus polishing of these CVD coated SiC samples was done to improve the starting surface roughness (polishing performed by RAPT Inc). The resultant surface roughness values for both samples are also shown in table 1. Vendor Original Surface Roughness (nm) Final surface Roughness after polishing (nm) Poco Graphite Inc. 1200 <100 Table 1. Surface roughness details of CVD coated SiC samples SCRATCHING USING A 5μm DIAMOND STYLUS TIP A 5μm radius diamond stylus was used for making scratches on the surface of CVD coated SiC. This diamond stylus was moved over a span of 5mm on the sample with continuously increasing loads using the load control mode in the tribometer. The loads were varied from 10 to 25 grams for the Poco Graphite sample. RESULTS FROM SCRATCHING USING A 5μm TIP The maximum ductile scratch depths that the Poco Graphite sample achieved were 700nm, which can be seen in figure 1. It is revealed from the peaks seen in the X-profile (upper right image) of the Wyko images that depths beyond this would be brittle, as the material is seen breaking up and moving out of the surface. The Y-profile (lower right image) is taken in the ductile region of the scratch, before the DBT is reached. There are two areas on the surface of a sample where ductile to brittle transition can take place. One is in front of the tool (at lower depths) and one is behind the tool (at higher depths). The DBT depth (approx. 700 nm) shown in this sample certainly is behind the tool as the material is pushed up above the surface-in the wake of the tool. The large peak (encircled region) seen in the Xprofile of figure 1 suggest the occurrence of brittle fracture in this region of the scratch. If the scratch is ductile there is usually material piled up on both sides of the scratch (as shown in fig 1, Y-Profile) unlike brittle fracture where the material moves above the surface of the sample unevenly. The forces (normal and tangential) were also measured along with the acoustic emission data using the tribometer. These data are shown in figure 2. The AE sensor was mounted on the tool holder away from the contact point of the diamond stylus and the surface of the sample. Hence it was difficult for the AE sensor to pick up signals from the brittle fracture. The forces shown are not for the entire scratch. The force plot exactly corresponds to the region of the scratch shown in figure 1. It is difficult to determine any difference in force values as the brittle fracture occurred behind the tool. Typically fracture behind the tool is not seen in the force plots, as it is outside the sensor’s measurement loop. To measure the force impulse from a major fracture event behind the tool usually requires additional sensitivity and/or instrumentation. Figure 1: Wyko image showing DBT depth behind the tool for Poco Graphite sample using 5μm diamond stylus Force plot for DBT depth