Micro-Scratch Testing and Simulations for Adhesion Characterizations of Diamond-coated Tools

Micro-Scratch Testing and Simulations for Adhesion Characterizations of Diamond-coated Tools Ping Lu Mechanical Engineering Department The University of Alabama Tuscaloosa, Alabama, USA Xingcheng Xiao and Michael Lukitsch Research & Development Center General Motors Corporation Warren, Michigan, USA Kevin Chou Mechanical Engineering Department The University of Alabama Tuscaloosa, Alabama, USA ABSTRACT In this study, micro-scratch tests were conducted on a diamond-coated tungsten-carbide substrate to investigate the coating adhesion. High intensity Acoustic Emission (AE) signals were detected once the coating delamination initiated during the scratch test. It has also been found that the tangential force increased gradually with the normal force, but varied significantly when the critical load of coating delamination was reached. A finite element (FE) model with a cohesive-zone interface was developed to simulate the scratch process and the coating delamination phenomena. The preliminary results indicate that it is feasible to use the FE combined with scratch tests to evaluate the coating interface characteristics. KEYWORDS Adhesion, Cohesive zone interface, Diamond-coated tool, Delamination, Finite element, Scratch testing. INTRODUCTION Applying hard coatings such as diamond on cutting tools is an ideal approach for enhancing tool lifetime and improving the machining quality due to diamond s high hardness/strength, low friction coefficient, and chemical stability, etc. Diamond-coated tools excel in cost and flexibility to various tool geometry and size comparing to synthetic polycrystalline diamond (PCD) tools (Hu, 2007), which are also commonly used in the manufacturing industry. Thus, diamond-coated cutting tools have a potential to replace costly polycrystalline diamond tools. However, in tooling applications, diamond coating delamination remains the primary wear mode that often result in catastrophic tool failures (Lu, 2009, Qin, 2009). Coating delamination is due to insufficient adhesion between the coating and substrate. Many approaches have been developed to enhance the coating adhesion such as etching cobalt (Co) phase off from the tool surface before diamond depositions. In addition, reliable experimental methods to quantitatively characterize the coating adhesion are critical to compare and further optimize different interface approaches. There are several methods used to examine the adhesion of coatings (Perry, 1983). For example, Bouzakis et al. applied an inclined impact test to evaluate the cohesion and adhesion properties of thin coatings (Bouzakis, 2004). Scratch testing is one of the most practical means of evaluating the adhesion of a hard, thin coating on a substrate (Bull, 1992, Ollendorf, 1999), since it is reliable, simple to perform and no special specimen shape or preparation are required. Adhesion is measured when a critical load is reached at which coating failure occurs. Provided that the failure is adhesive, this critical normal load is taken as a measure of the coating substrate adhesion, or the work of adhesion is derived from the

critical normal load (Jaworski, 2008, Nakao, 2007). Figure 1 is a schematic illustrating a scratch test on a coated specimen. During scratch testing, a spherical indenter tip slides over the surface of the coating to generate a groove under incremental or constant normal loads. The tangential force is measured during the test and the morphology of the scratches can be observed simultaneously or afterwards. When the mean compressive stress over an area in the coating exceeds a critical value, the coating detaches from the substrate to lower the elastic energy stored in the coating (Liu, 2009). The work of adhesion at the interface between the coating and substrate is equal to the energy release rate from coating detachment and this rate is a function of the mean compressive coating stress over the detaching area at the instant of detachment. Thus, the critical mean compressive coating stress responsible for the detachment could be a measure of coating substrate adhesion. On the other hand, diamond coatings are very brittle. While a coating can withstand compressive stresses induced by the indenter to a certain extent, it may fracture if a high tensile or shear stress field is induced simultaneously, in particular, at the interface such as delamination (Xie, 2002). 6% Co, with coating thickness around 4 µm, provided by University of South Florida. The surface roughness of the samples was about 0.38 µm Ra. A Micro-scratch tester from CSM Instruments, model Micro-Combi, was used for the experiments at the room temperature. Testing was conducted in typical laboratory temperature/humidity environments, The diamond indenter with a tip radius of 50 µm was utilized. The scratch speed was 2 mm/min with the progressive loading method. The scratch length for each test was 5 mm. During the scratch test, tangential forces, acoustic emission (AE) signals, and the depth of the scratch were acquired. A KEYENCE digital microscope (VHX- 600X) was used to observe the scratch marks and delamination after the test. In addition, a white-light interferometer (WLI) was used to acquire the morphology of scratch grooves. To determine the critical load for the tested diamond-coated tool, scratch tests were carried out with a progressive load, maximum normal force of 10 N, 15 N, 20 N, 25 N and 30 N. Figure 2 below shows the overall images of 5 scratch grooves at the corresponding load (1: 10 N, 2: 15 N, and 3: 20 N, etc.) Figure 1. A schematic illustrating a scratch test on a coated specimen (http://www.pvd-coatings.co.uk/scratch-adhesiontester.htm). This study aims at better understanding the adhesion of diamond-coated carbide tools by micro-scratch testing; critical load for coating delamination and the corresponding process signals were focused. In addition, a finite element (FE) model was developed to simulate the scratch process, with emphasis on the interface behavior, and to evaluate the adhesion between the coating and substrate using interface characteristics, which can be further applied in studies at different loading conditions such as cutting. EXPERIMENTAL DETAILS The specimen was a diamond-coated tool. The substrate is co-cemented tungsten carbide (WC-Co), submicron grains, Figure 2. Digital microscopic images of scratch grooves on the sample. RESULTS AND DISCUSSION Figure 3 shows the AE signal and tangential force (Ft) vs. the applied normal load (Fn) during the scratch test with a progressive load of maximum 10 N. From the figure, it is observed that the tangential force increases smoothly when the normal force is less than 5.4 N, but varies considerably once the load exceeds 5.4 N. It is also observed that an abrupt amplitude increase of AE signals (Spot 1) exists at the load around 5.4 N, followed by a series of continuous high-amplitude AE peaks. This implies coating delamination initiation occurred around 5.4 N.

Spot1 Figure 3. Acoustic emission (AE) and tangential force (Ft) vs. normal load (Fn) for a maximum load of 10 N. After testing, the scratched groove was observed in the digital microscope at 500X. From the relation between the load and the distance, the location corresponding to the high force variation and sudden intensity increasing (Spot 1) can be examined and verified whether coating delamination has initiated at that point. Figure 4a shows the digital microscopic image at Spot 1 and the corresponding normal load is around 5.4 N. It can be clearly confirmed that coating delamination has initiated at such a force, clearly exposing the substrate layer of WC, near Spot 1. Figure 4b shows the digital microscopic image at the end of scratch test. It is shown that coating delamination continued, once initiated, to the end of the final load, with a comparable delamination width. Figure 4. The digital microscopic images of Spot 1 around 5.4 N, and the end of the scratch for scratch test under maximum load of 10 N. For the conditions of different final normal loads, the tangential force/ae data show similar behaviors as the 10 N load, i.e., gradual increasing of tangential force and AE signals followed by an abrupt jump, and followed by large variations of force/ae magnitudes. For example, Figure 5 shows the AE signals and tangential force (Ft) vs. the normal load (Fn) during the scratch with a 30 N final load.

Spot1 Figure 5. Acoustic emission (AE) and tangential force (Ft) vs. normal load (Fn) for a maximum scratch load of 30 N. Further, using the approach described earlier, the corresponding critical load of delamination for each condition can be estimated. It is found that the critical load of delamination initiation is around 4 to 6 N for all tested loads, indicating that the delamination critical load is not sensitive to the load rates currently applied. Figure 6a shows the digital microscopic image at Spot 1 (for 30 N load scratching) with the normal load around 3.9 N. It is clearly seen that coating delamination has been initiated at this point (near Spot 1), an area of coating layer detached from the WC substrate. Figure 6b shows the digital microscopic image at the end of scratch test (maximum load 30 N). It is also noted that coating delamination continues spreading at the end of the scratches, and the delamination width becomes larger compared to at the beginning of delamination. Therefore, the coating delamination becomes more severe with the increased normal load. From the scratch test results, the evaluated diamond-coated tool with a coating thickness of 4 µm has a critical load for delamination in the range of 4 to 6 N. Figure 6. Digital microscopic images of Spot 1 around 3.9 N, and the end of the scratch (30 N load). The white-light interferometer has also been used to obtain quantitative information of the scratched geometry. Figure 7 below shows a 3D surface contour image of part of the scratched track, at ~3 mm distance from beginning (20 N maximum load). It clearly shows coating delamination propagated outward. Figure 8 is the 2D profile analysis of the scratched mark at the same location. It is estimated that the delamination width is about 88 µm and the depth of the track is about 7 µm. Figure 7. WLI image (3D) at 3 mm scratch distance for 20 N maximum load, showing coating delamination.

Figure 8. 2D scratch track profile analysis at 3 mm scratch distance for 20 N maximum load. SCRATCH PROCESS SIMULATIONS A 3D FE model has been developed to study the microscratch test by using ABAQUS 6.8. A half model is used because of the symmetry of the problem. The geometric model of the specimen shown in Figure 9a is 3000 µm in length and 200 µm in width. The thickness of the diamond coating and the carbide substrate is 4 µm and 60 µm, respectively. The diamond indenter is assumed to be rigid with a tip radius of 50 µm. Figure 9b shows the details of the model around the indenter. The smallest mesh size underneath the indenter is 0.6 µm for the coating and 5.0 µm for the substrate. Further, a cohesive layer with zero thickness was included to model the mechanical behavior of the interface between the coating and the substrate. The materials of the specimen, both the coating and the substrate, were modeled as an elastoplastic solid with isotropic hardening. According to the Ramberg-Osgood law, the stress-strain relation can be described as, where, and the strain-hardening exponent is generally in the range of 0 to 0.5 (Xu, 2005). The yield strength of WC-6wt%Co is in the range of 1.45 to 5.76 GP (Renaud, 2009, Mittal, 2001), and the yield strength of the WC-Co is assumed as 3.605 GPa in this study. Other properties of WC are identical with those in a previous study (Renaud, 2009). The yield strength and hardening exponent of the diamond can be obtained by the method proposed by Suresh (Giannakopoulos, 1999, Ramamurty, 1999, Venkatesh, 2000) based on the experimental data of nanonindentation on diamond (Chowdhury, 2004). In this study, the Young s modulus of the coating is 1200 GPa, other properties of the coating and substrate are listed in Table 1. The friction coefficient was set as 0.1. The cohesive zone properties such as the strength and characteristic length are derived from a previous study (Hu, 2008) with adjusted property values: 543 MPa maximum strength, 0.43 µm characteristic length and 117 J/m 2 fracture energy density, to approximate the critical load observed from the experiments.

Coating Indenter substrate Cohesive zone Figure 9. 3D FEA model for the scratch test: overall model, and details around the indenter. Table 1. Properties of the substrate and coating for the diamond-coated cutting tools Part Material Young s modulus/gpa Poisson s ratio Yielding strength/gpa Hardening exponent Substrate WC-6 wt. % Co 619.5 0.24 3.605 0.244 Coating Diamond 1200 0.07 26.7 0.23 Figure 10 shows the normal force and tangential force development along with the indenter location from the scratch simulation. The spherical indenter first moves downward in contact with the coating, then slides along the surface of the coating, at a speed of 500 µm/s, with a linearly increased depth, and the sliding distance is 500 µm. At the end, the slider moves upward until fully unloaded. Figure 10. Normal and tangential forces at different locations for the maximum load of 14 N. Figure 11 shows the maximum principal stress and the equivalent plastic strain in the specimen, around the indenter, during the initial sliding period, corresponding to around 2 N of the normal force. It can be seen that high compressive stresses, around 6 GPa, occur near the indenter and the maximum equivalent strain is 0.024. In addition, there is a small area with tensile stresses. As to be shown later, no delamination has been generated at this location. Figure 12 shows the maximum principal stress and the equivalent plastic strain in the specimen at the onset of coating delamination. It can be observed that high compressive stresses, around 20 GPa, occurred in the coating near the indenter and the maximum equivalent strain is 0.082. Moreover, high tensile stresses occur in the wake of the slider. The corresponding load at this location is about 4 N. And the delamination has just initiated, to be shown later. Figure 13 shows the maximum principal stress and the equivalent plastic strain in the specimen at the end of the sliding period. It can be seen that compressive stresses are as high as around 25 GPa, occurred in the coating near the indenter and the maximum equivalent strain is 0.34.

Figure 12. Maximum principal stress (unit: 1000 GPa) and Equivalent plastic strain at the onset of delamination. Figure 11. Maximum principal stress (unit: 1000 GPa) and Equivalent plastic strain at the initial scratching. Figure 13. Maximum principal stress (unit: 1000 GPa) and Equivalent plastic strain at the end of scratching.

Figure 14 illustrates the normal stress distribution in the cohesive-zone layer at different times: initial scratch, onset of delamination and (c) end of scratch. At the initial period, though the normal stress of the cohesive zone has reached the cohesive zone strength (543 MPa), the damage evolution has not resulted in the cohesive zone failure. Thus, the interface is still intact. On the other hand, at around 4 N of normal load, the cohesive-zone layer has shown some damage initiated; the cohesive zone normal separation has reached its characteristic length (0.43 µm). Figure 14c shows a large delamination area at the end of scratch testing. The coating delamination crack has spread in both the sliding and transverse directions. (c) Figure 14. Cohesive zone stress distribution (unit: 1000 GPa) at the interface: initial scratch, onset of delamination, and (c) end of scratch. Figure 15 further shows the cohesive-zone normal stress vs. normal separation traced along the scratch time at two different locations. Figure 15a is for a location where the interface remains attached. The normal stress is first compressive when the slider is approached to that location and quickly changes to tensile and reaches the interface strength. Then, the stress is reduced in a short time, but with the normal separation increased. Finally, as the slider moves away from this location, the stress is reduced, also the separation, but does not return to zero at the end. On the other hand, Figure 15b is for a location where the interface has been delaminated. The normal stress vs. separation curve follows the typical traction-separation behavior and the normal separation has reached the maximum separation and results in the interface failure. In a previous study of

diamond-coated carbides (Hu, 2008), the ideal cohesive zone properties (as an upper bound) were estimated from the substrate (WC) toughness. For the current study, note that the cohesive-zone properties have been modified (1/3 of the upper-bound fracture energy) to approximate the critical load measured from the experiments. Therefore, the estimated cohesive-zone characteristics may be used to indicate the interface capacity related to the delamination behaviors. During the scratch tests, the normal force, the tangential force, the acoustic emission signals and the penetration depth were acquired. After scratch tests, the scratch marks were also observed in a digital microscope and analyzed by a white-light interferometer. Moreover, to characterize the interface mechanical behavior, an FE model including the interface cohesive zone was developed to simulate the scratch process. The results are summarized as the following: (1) Coating delamination can be clearly detected by AE signals. It was observed that the abrupt AE peak jumps followed by several continuous AE high-amplitude peaks are associated with coating delamination. The tangential force increases smoothly with the normal force before the initiation of coating delamination, but, varies considerably once coating delamination initiated. Therefore, tangential force may also be used to monitor the coating delamination during scratch tests. (2) The width of coating delamination would increase with the increased loads after delamination initiated, this is confirmed by the scratch images under digital microscope, which show that the coating delamination becomes more severe with the increased load. (3) The critical load for the tested diamond-coated WC tools was in the range of 4 to 6 N, which has been confirmed by repeated tests. (4) The simulation results indicate that it is feasible to use the FE model combined with scratch tests to evaluate the coating interface characteristics. ACKNOWLEDGEMENTS This study was conducted under an NSF-funded joint project (CMMI 0928627) - GOALI/Collaborative Research: Interface Engineered Diamond Coatings for Dry Machining, between The University of Alabama, General Motors and University of South Florida. Figure 15. Normal stress vs. separation behavior along the scratch time: location without delamination, and location with delamination. Note that, it is known that during machining, the temperature will be substantially higher than the room temperature testing conducted here. The temperature increases may affect coating properties and the deposition thermal stress (Bouzakis, 2010). However, as the first step, this study attempted to establish the room-temperature adhesion data for general evaluation purposes. CONCLUSIONS Scratch tests of diamond coatings deposited on a WC substrate have been carried out using a micro-scratch tester. REFERENCES Bouzakis, K. -D., A. Asimakopoulos, N. Michailidis, S. Kombogiannis, G. Maliaris, G. Giannopoulos, E. Pavlidou, G. Erkens, (2004), The inclined impact test, an efficient method to characterize coatings' cohesion and adhesion properties. Thin Solid Films, Vol. 469-470, 254-262. Bouzakis, K. -D., M. Pappa, G. Skordaris, E. Bouzakis, S. Gerardis, (2010), Correlation between PVD coating strength properties and impact resistance at ambient and elevated temperatures. Surface and Coatings Technology, Vol. 205, 1481-1485. Bull SJ in: Gissler W, Jehn HA (Eds.) (1992) Advanced Techniques for Surface Engineering, ECSC, EEC, EAEC, 31 68.

Chowdhury S, de Barra E, Laugier MT (2004) Study of mechanical properties of CVD diamond on SiC substrates. Diamond and Related Materials 13, 1625-1631. Giannakopoulos AE, Suresh S (1999) Determination of elastoplastic properties by instrumented sharp indentation. Scripta Materialia 40, 1191-1198. Hu J, Chou YK, Thompson RG (2007) On Stress Analysis of Diamond Coating Cutting Tools. Transactions of NAMRI/SME 35, 177-184. Hu J, Chou YK, Thompson RG (2008) Cohesive Zone Effects on Coating Failure Evaluations of Diamond Coated Cutting Tools., Surface and Coatings Technology 203, 730-735. Jaworski R, Pawlowski L, Roudet F, Kozerski S, Petit F (2008) Characterization of mechanical properties of suspension plasma sprayed TiO2 coatings using scratch test. Surface and Coatings Technology 202, 2644-2653. Renaud A, Hu J, Qin F, Chou YK (2009) Numerical Simulations of 3D Tool Geometry Effects on Deposition Residual Stresses in Diamond Coated Cutting Tools. International Journal of Mechatronics and Manufacturing Systems 2, 490-502. Xie Y, Hawthorne HM (2002) Effect of contact geometry on the failure modes of thin coatings in the scratch adhesion test. Surface and Coatings Technology 155, 121-129. Xu Z, Li X (2005) Influence of equi-biaxial residual stress on unloading behavior of nanoindentation. Acta Materialia 53, 1913-1919. Venkatesh TA, Van Vliet KJ, Giannakopoulos AE, Suresh S (2000) Determination of elasto-plastic properties by instrumented sharp indentation: guidelines for property extraction. Scripta Materialia 42, 833-839. Liu FX, Yang FQ, Gao YF, Jiang WH, Guan YF, Rack PD, Sergic O, Liaw PK (2009) Micro-scratch study of a magnetron-sputtered Zr-based metallic-glass film. Surface and Coatings Technology 203, 3480-3484. Lu P, Chou YK, Thompson RG (2009) AE signal evolutions in machining by diamond coated tools. Proceedings of the 2009 International Manufacturing Science and Engineering Conference, MSEC2009-84372. Mittal KL (2001) Adhesion aspects of thin films V.S.P. Intl Science, Netheland. Nakao S, Kim J, Choi J, Miyagawa S, Miyagawa Y, Ikeyama M (2007) Micro-scratch test of DLC films on Si substrates prepared by bipolar-type plasma based ion implantation. Surface and Coatings Technology 201, 8334-8338. Ollendorf H, Schneider D (1999) A comparative study of adhesion test methods for hard coatings. Surface Coatings Technology 113, 86-102. Perry AJ (1983) Scratch adhesion testing of hard coatings. Thin Solid Films 107, 167-180. Qin F, Hu J, Chou YK, and Thompson RG (2009) Delamination wear of nano-diamond coated cutting tools in composite machining. Wear 267, 991-995. Ramamurty U, Sridhar S, Giannakopoulos AE, Suresh S (1999) An experimental study of spherical indentation of piezoelectric materials. Acta Materialia 47, 2417.