OARS - Open Access Rewards System
DOI : 10.2240/azojomo0284

Abrasive Grain Efficiency and Surface Roughness for Machining Ceramics by Newly Developed Cup-Type Diamond-Grinding-Wheels

TienDong Nguyen, Koji Matsumaru, Masakazu Takatsu and Kozo Ishizaki

Copyright AD-TECH; licensee AZoM.com Pty Ltd.
This is an AZo Open Access Rewards System (AZo-OARS) article distributed under the terms of the AZo-OARS https://www.azom.com/oars.asp which permits unrestricted use provided the original work is properly cited but is limited to non-commercial distribution and reproduction.
AZojomo (ISSN 1833-122X) Volume 6 November 2010

Topics Covered

Abstract
Keywords
Introduction
Experimental
     Newly Developed Cup Type Diamond Grinding Wheel with Hexagonal Structure
     Grinding Procedures
Results
Discussion
Conclusions
Acknowledgements
References
Contact Details

Abstract

Newly developed cup-type diamond-grinding-wheels with hexagonal pattern were used to grind a hard-to-machine ceramic material, which was represented by single crystal sapphire in the present work. The amount of diamond grains on the wheel surface were controlled by regulating the size of hexagonal shape and its edge thickness. Grinding stone ratio, R is defined as the ratio between the hexagonal edge area containing abrasive grains and the total area of the wheel surface. Four kinds of hexagonal grinding wheels with different R (12.9%, 19.0%, 24.9% and 36.0%) and a conventional wheel (R: 100%) were used. The number of abrasive grains that pass through a unit length of sample surface for each grinding pass, Ng was calculated to evaluate the efficiency of abrasive grains by wheel rotation speed, table feeding speed, and R. Surface roughness of ground sample was employed to determine the abrasive grain efficiency. When Ng increases, surface roughness becomes smaller, i.e., smoother surfaces. Surfaces ground by the conventional wheel are rougher than those by using newly developed hexagonal grinding wheels in spite of the larger Ng. Surface roughness data formed one curve for all hexagonal wheels in a graph of surface roughness versus Ng, and another curve for the conventional wheel. The results indicate that effective working abrasive grains in hexagonal wheels are about 5 times higher than those of a conventional wheel. In other words, only 1/5 of abrasive grains work effectively in the conventional wheel. Hexagonal structure shows advantage in the effective usage of abrasive grains in comparison with conventional structure.

Keywords

Hexagonal Structure, Cup-type Diamond Grinding Wheel, Sapphire, Ceramics, Abrasive-grain Efficiency, Surface Roughness

Introduction

Recently, advanced ceramic materials, e.g., Si3N4, SiC, or sintered Al2O3 have been applied in many industrial fields such as electronics and automobiles. These advanced ceramics, however, are difficult to machine due to their hardness and brittleness, thus machining costs constitute 80% or more of total component costs [1]. Single crystal sapphire, which has been used as a substrate for white light emitting diodes (LED), was selected to represent a hard-to-machine ceramic material in the present work. Sapphire thickness and its surface roughness are two important factors in LED fabrication process. Therefore, grinding process using diamond-grinding-wheels is widely used in order to obtain a smooth surface and precise thickness of sapphire substrates.

Kim et al. developed a regulated-force-feeding (RFF) grinding system for a cup-type grinding wheel in order to obtain a smooth sample surface with small sub-surface damages. In this advanced machining, table feeding force is kept constant by air cylinder system. On the other hand, a conventional grinding machine has a constant table feeding speed. The values of surface roughness of silicon wafers and sintered Al2O3 were almost constant for any cutting depth by using RFF grinding system, and increased as cutting depth increased for a regular grinding system of constant feeding speed [2, 3].

In conventional grinding process for ceramic materials, a smooth surface is obtained by controlling grinding parameters such as grinding wheel speed, cutting depth, and table feeding speed. Many researchers reported that ground surfaces became smoother by high wheel speed, low cutting depth or low table feeding speed. These researches used a straight-type grinding wheel, and modeled the kinematics of a single-grain-cutting-edge in order to obtain the relationship between chip cross-sectional area, scallop height or grit depth of cuts to surface roughness [4-9]. However, abrasive grains distribute randomly with different protrusion heights on the wheel surface, and may cause difficulty to obtain smoother surfaces. A cup-type wheel can grind more effectively to obtain smoother surface than a straight grinding wheel. There is no literature reporting the modeling to estimate the relation between grinding parameters and ground surfaces as far as we know. The modeling of a single-grain-cutting-edge can not be used for cup-type wheels, because controlling factors of surface roughness, e.g., chip cross-sectional area or scallop height are totally different from the case of straight-type grinding wheels. These problems have never been reported by any researches. According to the previous researches, smoother surface can be obtained by increasing rotation speed or decreasing cutting depth [4-9], but wheel rotation speeds or cutting depth have the limit depending on ground samples or grinding machines.

In this work, cup-type newly developed diamond-grinding-wheels with various amounts of abrasive grain on wheel surfaces are used to grind sapphire. The effects of abrasive grains and surface roughness of ground sample are evaluated. This paper reveals a new mechanism of grinding process by the proposed diamond-grinding-wheels.

Experimental

Newly Developed Cup Type Diamond Grinding Wheel with Hexagonal Structure

Diamond grinding wheels are newly developed with hexagonal pattern, which contain abrasive diamond grains on hexagon edges, and filled up inside of the hexagons by green carborundum porous material without diamond grains [10]. These wheels are characterized by hexagonal geometrical factors: size of hexagon, x, and width of hexagon edges containing diamond grains, w, as shown in Fig. 1. All the wheels have outer diameter of 250 mm and inner diameter of 80 mm. Grinding stone ratio, R is defined as the ratio between the hexagonal edge area containing abrasive grains and the total area of wheels. Four grinding wheels with different R (12.9, 19.0, 24.9 and 36.0%) and a conventional wheel with R 100% were used as shown in Table. 1.

Figure 1. Hexagonal diamond grinding wheel (outer diameter: 250 mm, inner diameter 80 mm) characterized by hexagonal geometrical factors: hexagon size, x, and an edge width, w. Diamond abrasive grains are placed only within the edge width.

Table 1. Grinding stone ratios, R for hexagonal wheels and a conventional wheel.

Hexagonal wheels, R
w/mm
x/mm
1.0 2.0
10 19.0% 36.0%
15 12.9% 24.9%
Conventional wheel, R 100%

Grinding Procedures

Fig. 2 shows the schematic diagram of the regulated-force-feeding (RFF) grinding system [2]. The grinding table was controlled by an air-cylinder system in order to regulate a constant table feeding force. In this work, the table feeding force was kept constant at 3.8 N. A vacuum vise was made by porous ceramics (NanoTEM Co. Ltd., Nagaoka, Japan) and was machined in order to achieve a perfect parallel surface between the vacuum vise and the working surface of grinding wheels. Samples were single crystal sapphire substrate with dimensions of φ50.8 x 0.45 mm, and the surface ground was C-plane {0001}. A sample was placed on vacuum vise so that the orientation flat <1100> was perpendicular to the feeding direction, i.e., the feeding direction is <1200> (normal to A-plane). Conventional and hexagonal diamond grinding wheels (#200, vitrified bond 20 weight %, NanoTEM Co. Ltd., Nagaoka, Japan) were used. Before each experiment, grinding wheels were balanced at 500 rpm using a dynamic balancing instrument in order to reduce vibration. After balancing, grinding wheels were trued by silicon carbide grinding stone (mesh no. 120) with grinding conditions of 10 μm cutting depth, 500 rpm wheel rotation speed in order to obtain flatness on the wheel surface. Then, grinding wheels were dressed by silicon carbide grinding stone (mesh no. 120) in order to sharpen abrasive grains under conditions of 10 μm cutting depth, 500 rpm wheel rotation speed until wheel thickness reduced about 5 μm.

Grinding wheel rotation speeds, n were 500, 1500 or 3000 rpm, cutting depth was 10 μm per pass, and total cutting depth was 200 μm. The wheels were re-dressed before each grinding experiment. The coolant water was sprayed into the contact zone between grinding wheel and sample during truing, dressing, and grinding process with 0.5 lmin-1 flow rate. Grinding conditions are listed in Table. 2.

Figure 2. Schematic diagram of the regulated-force-feeding (RFF) machining system. In this RFF grinding machine, table feeding force was kept constant by air cylinder system instead of keeping a constant Sf as in a conventional machine.

Table 2. Specifications of grinding wheels, grinding conditions and sample.

Grinding wheels Grinding conditions Sample
Diameters:
- Outer, DO 250 mm
- Inner, DI 80 mm
Grain size (#200), d 74 μm
Bonding material Vitrified
 
 
Rotation speed, n/ rpm 500, 1500, 3000
Cutting depth, 10 μm
Total cutting depth 200 μm
Constant table feeding force 3.8 N
Coolant water, 0.5 lmin-1
 
 
Material Sapphire
Diameter, Ds 50.8 mm
Thickness 0.45 mm
Grinding surface:
- C-plane {0001}
Grinding direction:
- Normal to A-plane <1200>

During the process, grinding time for each grinding pass was recorded to obtain a table feeding speed, Sf. The sample surface roughness, Ra was measured by using a profile-meter (Surfcom3000A, Tokyo Seimitsu Co. Ltd., Tokyo, Japan). Effects of R, and n on Ra were evaluated.

Results

The RFF machining system keeps a constant table feeding force, therefore Sf is altered depending on grinding wheel conditions, e.g., abrasive grain sharpness n, and R. The value of Sf can be calculated as:

where DO, DI, DS and t are outer diameter of a grinding wheel, inner diameter of a grinding wheel, sample diameter and time required for each grinding pass, respectively.

Figs. 3 and 4 show Sf and Ra as function of R for different n, 500, 1500 and 3000 rpm, respectively. For hexagonal wheels, at a given n, Sf is reduced as R increases. However conventional wheel with R 100% has a higher Sf in comparison with those for hexagonal wheels. For all grinding wheels, Sf becomes faster as n increases. Ra decreases as n and R increase. However, at a given n, higher Ra, i. e., a rougher surface is obtained by conventional wheel.

Figure 3. Tabel feeding speed, Sf as function of R for different n, 500, 1500 and 3000 rpm.

Figure 4. Sample surface roughness, Ra as function of R for different n, 500, 1500 and 3000 rpm. The data for conventional wheel do not fall on the extrapolated line for hexagonal wheels.

Discussion

When R and n increase, Ra becomes smaller, i. e., smoother for hexagonal diamond grinding wheels. However, higher Ra is obtained by conventional wheel with R 100% than by hexagonal wheels. In order to find controlling factors of Ra on ground samples, number of abrasive grains passed through unit length of sample surface, Ng is evaluated.

In the experiment, a sample is ground under different conditions of n and Sf. A point on the sample surface moves along an Archimedes' spiral on wheel surface as shown in Fig. 5. In other words, a point 1 of the sample at the starting moves to point 2 in Fig. 5. The value of Ng is the total number of abrasive grains on that Archimedes' spiral with pitch p and is calculated as:

where LA is Archimedes' spiral length, and za abrasive grains in a unit area on a conventional wheel surface. The pitch p of an Archimedes' spiral is calculated as:

The Archimedes' spiral length is calculated as:

If abrasive grains are distributed randomly and form a face centered cubic structure, the fictitious lattice parameter, a of the grain in the imaginary unit cell is calculated as:

where d and gv are grain size and the abrasive grain volume fraction, respectively. Grain size, d and gv are 74 μm and 0.48 for the present work, respectively.

Using Eq. (5), a is about 120 μm, the distance between grains, gd is order of 100 μm. However, some abrasive grains may drop out from wheel surface during truing, dressing or grinding process, therefore the real grain distance should be about twice of 100 μm. Distances between grains on wheel surface were observed by SEM (Scanning Electron Microscope) as seen Fig. 6, and the observed value of gd was about 200 μm. The amount of abrasive grains in a unit area on conventional wheel surface, za was estimated about 30 /mm2.

Figure 5. The way to calculate total number of abrasive grains, Ng , that passed through a sample surface for each grinding pass. Sample is grinding under conditions of n and Sf. A point 1 of the sample surface at the starting moves to point 2 along an Archimedes' spiral with pitch p on wheel surface.

Figure 6. SEM micrograph of an edge of a hexagonal diamond grinding wheel surface (#200, vitrified bond 20 weight %, NanoTEM Co. Ltd., Nagaoka, Japan). Grain distance, gd was estimated by averaging 54 measurements and the value is about 200 μm.

Figs. 7 (a) and (b) show Ra versus Ng. When Ng increases, sample surfaces become smoother. Surfaces ground by conventional wheel are rougher than those by hexagonal wheels in spite the higher Ng. Ra data do not form one curve, but two different curves, i.e., Curve (I) for the conventional wheel, and Curve (II) for the hexagonal wheels as shown in Fig. 7 (a). The possible cause of this phenomenon is the different number of effective working abrasive grains between two kinds of grinding wheels. By multiplying one fifth to the value of Ng for the conventional wheel, Curve (I) of conventional wheel superposes on Curve (II) for hexagonal wheels as shown in Fig. 7 (b). In other words, only one fifth of abrasive grains work effectively in the conventional wheel.

Figure 7. Surface roughness, Ra versus number of abrasive grains that passed through sample surface, Ng for each grinding pass. In (a), raw data are shown for Curve (I) for the conventional wheel and Curve (II) for hexagonal wheels. In (b), Curve (I) of the conventional wheel is superposed on Curve (II) of hexagonal wheels by multiplying one fifth to the value of Ng for the conventional wheel.

Working abrasive grains are defined as those that act directly to ground surface. On a grinding wheel surface, abrasive grains distribute randomly with different protrusion heights. The smoother ground surface is obtained by larger number of working abrasive grains. Fig. 8 shows schematic representation of the grooves made by diamond grains. On a conventional wheel surface, there are many abrasive grains moving one after another grain and pass through ground sample. However, only grains located at different position of groove formed by a previous grain or grains with high enough protrusion can touch and form another groove to remove ground material. Other grains, which are located to go through the same groove formed by previous grains and/or have lower protrusion height, do not touch ground surface. Ng includes both working and non-working grains. In this work, Ra data formed one curve for all hexagonal wheels in spite of the different hexagonal size, and formed another curve for conventional wheel. Hence, it can be concluded that almost all the number of abrasive grains work effectively by using hexagonal wheels within the range of the hexagonal dimension for the present work, while a conventional wheel has 1/5 lower effective working abrasive grains. Hexagonal structure shows advantage of increasing effective working abrasive grains in comparison with conventional grinding wheels.

Figure 8. Schematic representation of the way how diamond grains touch sample. Ground groves are formed by cutting grains, and smoother surface is obtained by a larger number of cutting grains, i.e., working abrasive grains, which make more grooves on ground sample obtaining smoother surface.

Conclusions

In this work, sapphire is ground by conventional and newly developed hexagonal diamond grinding wheels. R is defined as the ratio between the hexagonal edge area containing abrasive grains and the total area of wheel surface. The effects of R on sample surface roughness are evaluated. The following can be concluded:

  1. Number of abrasive grains is a governing factor of surface roughness. It is easy to imagine having smoother surface with more grains to pass. The conventional grinding wheels, however, have a limit to increase the number of working abrasive grains. Hexagonal wheels can overcome the shortcoming of conventional wheels by distributing abrasive grains within edges of hexagon and characterized by size of hexagons.
  2. Only 1/5 of abrasive grains work effectively in the conventional wheel.
  3. It is possible to have smoother surface by hexagonal wheels in comparison with a conventional wheel by controlling structure of a grinding wheel.

Acknowledgements

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 and Promotion of Independent Research Environment for Young researchers of the Ministry of Education, Culture, Sport, Science and Technology and Japan Science and Technology Agency.

References

1. J. A. Kovach, P. J. Blau, S. Malkin, S. Srinivasan, B. Bandyopadhyay and K. Ziegler, “A Feasibility Investigation of High Speed, Low Damage Grinding Process for Advanced Ceramics” The 5th International Grinding Conf (SME), 1 (1993).
2. H. Kim, K. Matsumaru, A. Takata and K. Ishizaki, “Grinding Behavior of Silicon Wafer and Sintered Al2O3 by Constant Force Feeding Grinding System”, Adv. in Tech. of Mat. and Mat. Proc. J. (ATM), 5 [2] (2003) 50-53.
3. H. Kim, K. Matsumaru, A. Takata and K. Ishizaki, “Reduction of Ceramic Machining Defects by Regulated Force Feeding Grinding System”, Adv. in Tech. of Mat. and Mat. Proc. J. (ATM), 6 [2] (2004) 290-297.
4. G. Warnecke and U. Rosenberger, “Basic of Process Parameter Selection in Grinding of Advanced Ceramics”, Annals of the CIRP (College International pour la Recherche en Productique), 44 (1995).
5. J. E. Mayer and G. P. Fang, “Effects of Grinding Parameters on Surface Finishing of Ground Ceramics”, Annals of the CIRP (College International pour la Recherche en Productique), 44 (1995).
6. E. C. Bianchi, P. R. Aguiar, E. J. da Silva, C. E. da Silva and C. A. Fortulan, “Advanced Ceramics, Evaluation of the Ground Surface”, Ceramica, 49 (2003) 174-177.
7. S. Malkin, “Grinding Technology Theory and Applications Machining with Abrasives”, Ellis Horwood Limited Publication, Chichester, England (1989).
8. Milton C. Shaw, “Principles of Abrasive Processing”, Oxford Science Publications, USA (1996).
9. I. D. Marinescu, W.B. Rowe, B. Dimitrov and I. Inasaki, “Tribology of Abrasive Machining Process”, William Andrew Publication, 13 Eaton Avenue, Norwich, NY 13815 (2004).
10. K. Ishizaki and A. Takada, Patent No-156724, 15th June 1999, Japan.

Contact Details

Ashok Kumar
Department of Mechanical Engineering
University of South Florida
Tampa FL 33620; USA

Email: [email protected]

Sathyaharish Jeedigunta
Department of Electrical Engineering
University of South Florida
Tampa FL 33620; USA

I. Tarasovc and S. Ostapenkoc
Nanomaterials and Nanomanufacturing Research Center
University of South Florida
Tampa FL 33620; USA

This paper was also published in print form in "Advances in Technology of Materials and Materials Processing", 10[2] (2008) 95-100.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit