The usage of hard and brittle materials, such as engineering ceramics and optical glasses, has been significantly increased to accomplish downsizing, low energy-consumption, and new applications for electronic, optical, or magnetic devices. For the machining of such materials, diamond-grinding stones (GSs) are widely used. A diamond GS consists of diamond abrasive-grains, supported by a matrix material, and pores (sometimes no pores). In order to achieve high efficiency and precision in grinding, a metal matrix diamond GS is an effective tool due to high rigidity of the matrix. Contrarily, such kind of GSs require a surface conditioning process called dressing to remove a matrix material and to maintain effective cutting edges of abrasive-grains. Most commonly, a dressing process is mechanically performed, machining the surface of a GS by a dresser. This mechanical process is influenced by many parameters, such as types of a GS and dresser, dressing speed, lubricant, and so on, which give an operator difficulties to obtain the desired distribution of cutting edges on a GS surface. The authors’ group has developed a laser-dressing process , which has been used to dress fabricated porous cast-iron matrix diamond GSs [2,3]. The selective removal of a matrix material was achieved with a minimal damage on abrasive-grains [4,5]. The dressing plays an important role on the surface topography of GSs, e. g., an abrasive-grain protrusion height. It is crucial to control the surface topography by dressing for highly efficient and precise grinding. In this report, the laser dressing process is performed on a porous cast-iron matrix diamond GS, and an average abrasive-grain protrusion height on the dressed surface is measured. To show the efficiency of the laser dressing, the specific grinding energy for a laser dressed GS is evaluated after grinding a zirconia ceramic.
Porous cast-iron matrix diamond GSs (150-200 μm size diamond grain: 19 vol%, 5 μm size cast-iron: 56 vol%, and porosity: 26 vol%) were prepared by the same manner reported elsewhere . Dressing was conducted on the as-sintered surface of a GS, which had the maximum surface roughness, Ry, of approximately 10 μm. Laser dressing was made by Nd-YAG laser at irradiation conditions of 532nm-wavelength, 1 kHz-repetition, 0.01 MJm-2-fluence, and 0.05 mms-1 scanning speed in a laser irradiation system shown in Figure 1 (LGS).
As a comparison, a GS mechanically dressed by a vitrified matrix GC#180 stick-type dresser was also prepared (MGS). The dressing was performed under a water supply of 10 mm3s-1, a dressing speed of 6 ms-1, a dressing feed of 5 μm per pass, and a total dressed depth of 50 μm. The grinding surfaces of the prepared GSs were shaped in a rectangular of bxl=4 mm x 13 mm, where b and l are the width and length of the GS, respectively. Both GSs were examined through constant pressure grinding tests at 0.4 MPa and a grinding speed, V, of 10 ms-1, grinding yttrium-partially stabilized zirconia (Y-TZP; Young’s modulus: 211 GPa, bending strength: 897 MPa, Vickers hardness: 11.9 MPa, and fracture toughness: 7.2 MNm-3/2) for 10 min under a water supply of 10 mm3s-1. The schematic illustration of the grinding system is shown in Figure 2. Before grinding, the surface of the ground material was machined by SD600V (vitrified bonded diamond grinding stone, #600), ensuring a surface roughness within 10 μm in Ry.
Figure 1. The laser irradiation system used in the present study. A laser beam generated at a laser head in a diameter of 0.7 mm is irradiated to the surface of a sample GS through optical mirrors and lenses. A sample GS is operated by an automatic stage for laser-scanning. Laser fluence is estimated with the beam diameter and power monitored by a power meter.
Figure 2: Schematic of a reverse constant pressure grinding test system. A dressed GS is fixed on a holder and pressed onto a rotating cup-shaped ground material. The surface of the Y-TZP was machined by a vitrified matrix #600-diamond grinding stone (SD600V) before the grinding test. The origin of the coordination is at the center and on the ground surface of the rotating Y-TZP.
In the grinding test, the Y-TZP was in a cup-shape rotating and the GS sample was pressed onto it. A tangential grinding force, F, and the displacement, d, of the GS were detected by a load cell and eddy-current displacement sensor, respectively. For grindability evaluation, removal volume ratio, v, and specific grinding energy, e, were calculated by the following equations:
where t is grinding period. The coordination taken in the present report is indicated by z, r, and θ, corresponding to the depth, radius distance, and azimuthal angle of a cup-shaped GS, respectively, as shown in Figure 2.
Dressed and ground surfaces of the GSs were observed by a Confocal Laser-scanning Microscope (CLM), and the surface section area, A, was measured at each depth, z, of a GS surface. An observed view was adjusted to a rectangular area of 1.17 mm by 1.47 mm. On the rectangular GS surface of 4 mmx13 mm, five views were observed, which corresponds to 16% of the total area. The surface section area ratio, A/Ao, where Ao is a total surface area, as a function of z was obtained for the evaluation of abrasive-grain protrusion heights following the manner previously reported . The ground surface of the Y-TZP was also analyzed through the observation by CLM, focusing on a ground groove number in a unit length perpendicular to the grinding direction.
Results and Discussion
A series of contour maps of a laser dressed surface as a function of z is shown in Figure 3. The Dark part is A at each depth. When A appeared on abrasive-grains, the number of grains and the depth were noted. After the observations by CLM on both dressed and ground surfaces of tested grinding stones, grain density and surface section area ratios were obtained. Figure 4 shows the grain density, Ng, and A/Ao as a function of z and their changes after grinding. Solid marks and open marks indicate data from as-dressed surfaces and ground surface, respectively. The LGS showed higher Ng, compared to the MGS showing 30% lower Ng with the dislodgment after dressing. Besides, the LGS showed no grain dislodgment after grinding while Ng of the MGS decreased by 1 mm-2 after grinding. The A/Ao distribution for the LGS changed towards higher z, showing the matrix deformation after grinding, which could not be seen in that of the MGS. Figure 5 shows the difference of Ng or cutting edge density, ΔNg, and the difference of surface section area, Δ(A/Ao), as a function of z. ΔNg for the LGS showed a twice higher value than that for the MGS, which lost grains during the dressing. The ΔNg–z curve shows that the LGS has aligned grain distribution with respect to depth. The matrix surface of the MGS was flat and there was negligible difference through the grinding while the surface of the LGS matrix showed a 10 μm-shift of the maximum Δ(A/Ao) along z due to deformation.
Distributions of ΔNg and Δ(A/Ao) give average values, considering these as probability distributions. According to the Δ(A/Ao) distribution and ΔNg distribution, the median value of depth with respect to Δ(A/Ao) is taken as a base line, and the median of depth at ΔNg is considered as an average cutting edge depth. Thus, average abrasive-grain protrusion heights for the dressed LGS and the MGS are 12.9 μm and 43.7 μm, respectively. The protrusion height of the LGS changed to 25.2 μm after grinding due to the deformation of matrix material, and that of the MGS slightly decreased to 42.9 μm caused by grain dislodgment.
Figure 3. A series of contour maps of a laser dressed surface as a function of depth, z. A dark part, surface section area A, increases with z.
Figure 4. Distributions of the grain density, Ng, and the surface section area ratio, A/Ao, along depth, z. The LGS showed higher Ng and no grain dislodgment after grinding, compared to the MGS showing 30 % lower Ng with the dislodgment after grinding. The distribution A/Ao showed the matrix deformation after grinding for the LGS but not for the MGS.
Figure 5. Cutting edge density, ΔNg, and the difference of surface section area ratio, Δ(A/Ao), as a function of depth, z. The depth at the median of Δ(A/Ao) is taken as a base line and the depth at the median of the ΔNg distribution is considered as an average cutting edge depth. Thus, average abrasive-grain protrusion heights for as-dressed LGS and MGS are 12.9 μm and 43.7 μm, respectively. The protrusion height of the LGS changed to 25.2 μm after grinding due to the matrix deformation.
For the LGS, the initial average abrasive-grain protrusion height easily changes because of deformation of a matrix material, and therefore, the real protrusion height during grinding is higher than the initial value.
Surface profiles along r direction on Y-TZP surfaces ground by both the LGS and MGS were obtained and shown in Figure 6. By observations on the ground surfaces of Y-TZP, 9 mm-1 grooves for the LGS and 6 mm-1 grooves for MGS were confirmed in a unit length of r. The difference between these groove numbers was 30%, thus, there must be 30% higher number of effective grains on the LGS surface. This difference in groove numbers between the LGS and MGS relatively agrees with the difference in abrasive-grain density described in Figure 4 above. Through the constant pressure grinding of Y-TZP, v shown in Figure 7 and e in Figure 8 were obtained with respect to grinding lengths, L (L=l x t x the revolution of a cup-shaped Y-TZP). The value of v decreases with L in both cases of the LGS and MGS.
Figure 6. Surface profiles along a radius direction, r, (the coordination is shown in Figure 1) on Y-TZP surfaces ground by the LGS (a) and by MGS (b). The LGS formed grooves of 9 mm-1 while the MGS formed approximately 30% lower number of 6 mm-1. The difference between these numbers agrees with the difference of abrasive-grain density between the LGS and MGS, which were observed in as-dressed surfaces.
Figure 7: Removal volume rates, v, as a function of grinding length, L. The value of v decrease with L. In the beginning of grinding, v of the LGS is more than twice higher than that of the MGS.
Figure 8: Specific grinding energy, e, as a function of grinding length, L. The LGS shows constantly lower e than the MGS. The value of e for the MGS suddenly shows significantly high values at some points on L.
In the beginning of grinding, v of the LGS is more than twice higher than that of the MGS. Over L of 150 m, v of the LGS constantly higher than that of the MGS. Correspondingly, e increases with L, and e for the LGS shows a lower value than that for the MGS. In L of 150 to 200 m and at 300 m, the MGS showed significant high values of e. These values may indicate a grinding energy increment with abrasive-grain dislodgment. Throughout the grinding tests, the LGS showed energy increment with abrasive-grain dislodgment. Throughout the grinding tests, the LGS showed constantly higher grinding efficiency than that of the MGS.
As summary, the developed laser dressing technique could maintain the initial grain distribution of a GS without any grain dislodgment, and consequently, achieved lower specific grinding energy than the GS dressed conventional mechanical method.
In order to achieve highly efficient machining of ceramics by grinding stones, the laser dressing was applied on a porous cast-iron matrix diamond grinding stone (LGS), using the 2nd harmonic generation of pulsed Nd-YAG Laser. The LGS achieves lower specific grinding energy to machine partially stabilized zirconia than the mechanically dressed grinding stone (MGS) through 300 m constant-pressure-grinding tests. As it is confirmed by the surface observation, the grain distribution of the LGS is more aligned with respect to depth than that of the MGS because of the no abrasive-grain dislodgment. As a result, a higher number of ground grooves are formed on the ground material surface, compared to those of the MGS. Therefore, the LGS contain higher number of effective cutting edges and achieves lower specific grinding energy.
The authors wish to express their gratitude to the Japanese government for partially supporting this work through the 21st Century Center of Excellency (COE) Program of the Ministry of Education, Culture, Sports, Science and Technology. A part of this research was also supported by New Energy and Industrial Technology Development Department (NEDO), Japan.
1. K. Ishizaki and A. Takata, “Non-contact Dressing and Truing Method for Grinding Stones and Its Apparatus”, Japanese Patent JP11-285971.
2. H. Onishi, Y. Kondo, S. Yamamoto, A. Tsukuda, and K. Ishizaki, “Fabrication of Porous Cast-Iron Bonded Diamond Grinding Wheels and Their Evaluation to Grind Hard-to-Grind Ceramics”, J. of Ceram. Soc. of Jpn., 104  (1996) 610-613.
3. H. Tomino, A. Tsukuda, Y. Kondo and K. Ishizaki, “Influence of Porosity on Grinding Performance of Porous Cast-iron Bonded Diamond Grinding Wheel Made by Pulse Electric Current Sintering Method”, J. of Jpn. Soc. of Powd. and Powd. Metall., 46  (1999) 257-261.
4. K. Jodan, H. Funakoshi, K. Matsumaru and K. Ishizaki, “Laser Dressing Process of Porous Cast-Iron Bonded Diamond Grinding Wheels for Machining Ceramics”, Adv. in Tech. of Mater. Mater. Proc. J. (ATM), 2  (2000) 117-123.
5. H. Funakoshi, K. Jodan, K. Matsumaru, and K. Ishizaki, “Laser Dressing Process of Porous Cast-Iron Bonded Diamond Grinding Wheels for Machining Ceramics”, Interceram, 50  (2001) 466-469.
6. K. Jodan, K. Matsumaru, M. Nanko, and K. Ishizaki, “Precise Method to Estimate Abrasive-Grain Protrusion Heights for Grinding Stones”, ATM, 5  (2003) 25-30.
Kazuki Jodan, Koji Matsumaru and Kozo Ishizaki
Nagaoka Gijutsu-Kagaku Daigaku (Nagaoka University of Technology)
Nagaoka, Niigata, 940-2188
E-mail: [email protected]