Wear of the blade diamond tools in truing vitreous bond grinding wheels Part I. Wear measurement and results

Wear 250 (2001) 587 592 Wear of the blade diamond tools in truing vitreous bond grinding wheels Part I. Wear measurement and results Albert J. Shih a,, Jeffrey L. Akemon b a Department of Mechanical and Aerospace Engineering, North Carolina State University, P.O. Box 7910, Raleigh, NC 27695-7910, USA b Cummins Engine Company, Columbus, IN 47202-3005, USA Abstract The wear of stationary blade diamond tools used to generate a precise and intricate form on the vitreous bond grinding wheel is presented. Two types of blade tools made of rod and particle diamond were used. A method to measure the wear of the blade diamond tool in the m-scale range using the size difference of two parts ground before and after truing was introduced. Two sets of experiments with four truing feeds and four tool traverse speeds across the grinding wheel were conducted on the rod and particle blade diamond tools, respectively. Experimental results showed the wear rate of blade diamond tools was improved at higher truing feeds and traverse speeds due to the brittle fracture of the abrasive and vitreous bond. 2001 Elsevier Science B.V. All rights reserved. Keywords: Diamond wear; Wear measurement; Ceramic grinding 1. Introduction The wear of stationary blade diamond tools used to generate a precise and intricate form on a rotating vitreous bond grinding wheel is presented. As shown in Fig. 1(a), a stationary blade diamond tool is used to generate or true the form on a grinding wheel. In grinding terminology, truing is the process to generate the form on a grinding wheel. Fig. 1(b) shows that the grinding wheel, after truing, is plunged into a rotating part to generate the desired cylindrical form. To create intricate part features, such as the small corner radius at A and B in Fig. 1(a), the blade diamond tool needs to be thin, yet still wear resistant. Wear on the thin diamond blade tools generates form errors on the grinding wheel and, subsequently, on the ground part. Slow and steady wear of diamond tools is the key to achieving consistent wheel surface conditions and reliable grinding results. For truing the vitreous bond SiC wheel specially designed for ceramic grinding [1], the wear of thin diamond blade tools is very important in controlling the form accuracy and consistency of ground parts [2 5]. This study develops a method to measure the wear of the blade diamond tools at different truing feeds and traverse speeds and analyzes results of the wear rate. As shown in Fig. 2, rod and particle diamond blade tools are used in this study. The rod diamond blade tool, as shown in the upper side and front views in Fig. 2(b) and (c), has Corresponding author. Tel.: +1-919-515-5260; fax: +1-919-515-7968. E-mail address: ajshih@eos.ncsu.edu (A.J. Shih). several long, thin diamond rods aligned and embedded in the metal matrix. These diamond rods usually have a square cross-section. To manufacture the thin diamond rod, the chemical vapor deposition method was used to form a thin layer of diamond, which was then polished and sliced into the rod shape. The other type of tool used in this study is the particle diamond blade tool, as shown in the lower side and front views in Fig. 2(b) and (c). The fine diamond grit and binder material were mixed, packed, and sintered to form the desired thin blade shape. This thin diamond blade is also aligned and embedded in the metal matrix to support the blade during truing. The width of the diamond rod and the thickness of the particle diamond blade used in this study, w in Fig. 2(c), are about 0.3 mm. Compared to the particle diamond blade tool, which has been widely used in production grinding, the rod diamond blade tool is a relatively new technology for precision truing of the grinding wheel. One of the objectives of this study is to quantify the performance of this type of diamond tool. Werner and Minke [2] studied the wear of several stationary single-point polycrystalline and natural diamond tools for truing aluminum oxide wheels. The diamond tools used were relatively larger than the diamond tools used in this study and had definitive geometry. Shih [6] investigated the wear of rotary diamond truing of vitreous bond diamond wheels. Neither study explored the effects of truing speed and feed. Another objective of this study is to examine the effects of these two parameters on the wear of the diamond tools. 0043-1648/01/$ see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0043-1648(01)00610-X

588 A.J. Shih, J.L. Akemon / Wear 250 (2001) 587 592 Fig. 1. Form generation on a grinding wheel using the blade diamond tool: (a) truing the grinding wheel to the desired form using a blade diamond tool; (b) plunge grinding using a formed grinding wheel. In this paper, the development of a technique to accurately measure the wear of stationary diamond tools is first introduced. The experiment designed to quantify the effects of three truing parameters, i.e. feed, traverse speed, and type of diamond tool, is then presented. A wear parameter is defined. Experimental results of the diamond tool wear are discussed. 2. Wear measurement of blade diamond tools To generate a sub- m-scale precision form on the grinding wheel, the wear rate of blade diamond tools during truing must be slow. A method to accurately measure the wear of blade diamond tools is developed in this study. At first, a micrometer with 0.5 m resolution was used to measure the change in length of the diamond tool, l in Fig. 2(a) and (b). This direct measurement method failed to yield satisfactory, repeatable measurement of l due to the following reasons: 1. The surface on blade diamond tools in contact with the grinding wheel is rough due to the erosion of the metal bond around the diamond grits during truing. SEM pictures of worn blade diamond tools, to be presented in Part II of the paper, reveal such rough surface conditions, which make the direct distance measurement in the m-scale precision level unfeasible. 2. As shown in Fig. 2(a), due to the curvature of the grinding wheel, the surface on the blade diamond tools in contact with the grinding wheel is also curved. This, compounded with the random protrusion of diamond grits, further hindered the precision wear measurement. For both types of blade diamond tools, experimental results to be presented in Part II of the paper showed that the truing force is not uniform. There are intermittent spikes of high truing force in a series of truing passes with the same process parameters. This indicates that the diamond on the blade tools is micro-fracturing and a large number truing passes is required to get a representative value of the average wear rate of diamond tools. In this study, a large number of truing passes was conducted until the desired 20 30 m wear on blade diamond tools was reached. The method developed in this study to measure the 20 30 m level of wear on the diamond tool is to use a Computer Numerical Controlled (CNC) cylindrical grinding machine, which can consistently grind parts to the m-scale precision level. After the machine is warmed up, the wear of the blade diamond tool is equal to half of the difference in diameter of parts ground before and after truing. This is explained in the remainder of this section. Configurations of the machine for truing and grinding are shown in Fig. 3(a) and (b), respectively. This machine has two slides, X and Z, with 1 m resolution. The workhead spindle, which drives the rotating part, is fixed on the X-slide. The grinding wheel spindle is attached on the Z-slide. Directions of motion of the two slides are perpendicular to each other. To true a grinding wheel, as shown in Fig. 3(a), the X-slide is first positioned to set the overlap between the grinding wheel and blade diamond tool to δ, the truing feed. The X-slide then remains stationary and the Z-slide moves across the grinding wheel at a constant traverse speed, v d. A close-up view of the truing process and its parameters is shown in Fig. 4. Fig. 4(a) and (b) show the beginning and ending of a set of truing passes, respectively. The blade diamond tool moves across the grinding wheel in a specific direction, as shown in the trajectory in Fig. 4(b). A set of Fig. 2. The rod and particle diamond blade tool: (a) side view of the blade diamond tool truing the grinding wheel; (b) side view of rod and particle diamond tools and the curvature on used diamond tools; (c) front view of the diamond tool surface in contact with the grinding wheel.

A.J. Shih, J.L. Akemon / Wear 250 (2001) 587 592 589 Fig. 3. Configurations for truing and grinding in a cylindrical grinding machine: (a) blade diamond tool truing a grinding wheel with the truing feed, δ, and traverse speed, v d ; (b) plunge grinding a workpiece with the width of grinding, b, and feed rate, v f. truing process includes three parameters: N, the number of truing passes, δ, and v d. Fig. 4 also illustrates the diameter of the grinding wheel before and after truing, d s1 and d s2. The length of the blade diamond tool before and after truing is defined as l 1 and l 2, respectively. The wear of the blade diamond tool after a set of truing (N, δ, v d )is, where is equal to l 1 l 2. After truing the wheel, as shown in Fig. 3(b), the X and Z slides are positioned to grind a rotating workpiece. The Z-slide is moved to set a given width of grinding, b. Then, the Z-slide remains stationary and the X-slide plunges the grinding wheel into the workpiece at a constant feed rate, v f. Truing changes the size of a grinding wheel. After a set of truing passes, if there is no wear on the blade diamond tool, the diameter of the grinding wheel is reduced by 2Nδ. In order to maintain the same size on the ground part, the position of the X-slide at the end of the plunge grinding is advanced by a distance Nδ to compensate for the wheel size change. Most CNC grinding machines, including the one used in this study, perform this compensation automatically in the program. The wear of the diamond tool changes the size of parts ground before and after truing, which is the principle used in this study to measure the wear of blade diamond tools. The diameter of ground parts is measured in the sub- m-scale level using a laser micrometer. The d s2, the diameter of grinding wheel after truing, can be expressed as d s2 = d s1 2Nδ + 2 (1) The actual size of the grinding wheel after truing, d s2,is2 bigger than the ideal, no tool wear case. Due to the blade diamond tool wear, a part ground after truing is 2 smaller in diameter than the part ground before truing. An accurate grinding machine, which can consistently grind parts to the same size, is the key for the proposed method. A test was conducted to quantify the accuracy of the cylindrical grinding machine, model U80 built by UVA in Sweden, used for this study. Fig. 4 shows the size variation of 15 parts ground, without changing the set up, as the machine gradually warmed-up in a 4 h period. Slides are moving, spindles are running, and coolant is spraying on the wheel, spindles, and slides during the machine warm-up. The wheel wear is negligible in grinding the machinable plastic parts used in this test. The main source of error for Fig. 4. Detailed view of (a) the blade diamond tool truing of the grinding wheel with feed, δ, and traverse speed, v d ; (b) the wear of the blade diamond tool,, after a set of truing with N passes.

590 A.J. Shih, J.L. Akemon / Wear 250 (2001) 587 592 Fig. 5. The thermal error during grinding machine warm-up. the roughly 40 m reduction in part size is mainly due to the machine tool thermal error, i.e. the thermal expansion of grinding wheel, tooling, and various machine components during the machine warm-up cycle. Fig. 5 shows that, after 2 h of machine warm up, the error on part size variation is reduced to a 3 m band. This observation demonstrates that this grinding machine has the required accuracy to measure the blade diamond tool wear in the 20 30 m range. Fig. 6 summarizes the steps used to measure the wear of blade diamond tools. First, the machine is warmed up for at least 2 h. The grinding wheel is then trued to the spindle axis and a machinable plastic part is ground after truing. The size of the part is designated as d p1. The grinding wheel is then trued repeatedly under a given set of truing parameters (N, δ, and v d ) until the tool wear reaches the desired 20 30 m range. After truing, the second machinable plastic part is ground and its diameter is measured as d p2. The wear of the blade diamond tool,, is equal to (d p2 d p1 )/2. This concept has been proved in this study as a reliable and repeatable method for measuring the wear of the blade diamond tool. 3. Experimental setup and design Fig. 6. Steps to measure the blade diamond tool wear in a CNC grinding machine that automatically compensates the truing depth in plunge grinding. The key set up parameters for the grinding machine, truing process, grinding wheel, coolant, and workpiece are summarized in Table 1. Three parameters, truing feed, traverse speed, and the type of blade diamond tool, were varied in the test. Table 2 shows the matrix for the 16 truing tests, including four different truing feeds and four traverse speeds. The truing lead, s d,is the axial pitch of the diamond tool per wheel rotation [2]. Each test required an extensive number of truing passes to generate the 20 30 m tool wear. These 16 truing tests were repeated for the rod and particle diamond blade tools. In summary, 32 truing tests were conducted. The power change in the grinding wheel spindle was measured using a Hall effect power sensor [6,7]. The power signal is then used to calculate the tangential truing and grinding force for the truing and grinding process, respectively. The results of the truing and grinding forces will be discussed in Part II of the paper.

A.J. Shih, J.L. Akemon / Wear 250 (2001) 587 592 591 Table 1 Summary of the setup parameters for truing and grinding Machine Machine Spindle type Slide resolution Workhead speed Workhead rotational direction Radial plunge speed Truing Radial feed, δ Setup parameter #1 Traverse speed, v d Setup parameter #2 Diamond tool Setup parameter #3 UVA U80 CNC cylindrical grinding machine Ball bearing 1 m 1200 rpm Same as the grinding wheel 5.0 mm/min 1, 2, 5 or 10 m 13, 25, 50 or 125 mm/min Particle or rod diamond blade tool Grinding wheel Grinding spindle motor 3 kw (4 HP) Power sensor Load control PH-3A 460 V, 10 A, 0 10V dc Abrasive and bond type Vitreous bond SiC Balancer No Diameter and width 305 mm (12 in.) in diameter and 20 mm wide The rpm and surface speed 1800 rpm, 29 m/s (for 305 mm diameter wheel) Coolant Type Filtration Temperature control Capacity Flow rate Table 2 The 16 truing tests matrix Traverse speed (mm/min) Truing lead, s d ( m) Cincinnati Milacron, Cimtech 500, 5 6% concentration 5 m paper filter None 113 liter (30 gallon) 2.0 l/min Radial feed, δ ( m) 1 2 5 10 In this study, the volume of wear of the rod diamond is estimated by multiplying the tool wear ( ) by the total cross-sectional area of the diamond rods. The volume of wear of the particle diamond blade tool is difficult to quantify. Instead, the length of the wear on the blade diamond tool,, is used to define another wear parameter G t. G t = V w (3) The wear on the blade diamond tool reduces the value of V w from the ideal, no blade tool wear case. V w is calculated using the following formula: V w = h w π(d 2 s1 d2 s2 ) (4) where h w is the width of the grinding wheel. Accurate measurement of diameters d s1 and d s2 of a 300 diameter vitreous bond grinding wheel in the m-scale level is difficult. In this study, only d s1 is measured using a caliper with 25 m resolution. The d s2 is calculated from d s1 using Eq. (1). This makes the error in the measurement of the grinding wheel diameter (25 m in 300 mm range) not a sensitive factor in the calculation of G t. It can be seen in the formula for V w after substituting Eq. (1) into Eq. (4). V w = 4h w π[d s1 (Nδ ) (Nδ ) 2 ] (5) The results of G d for the rod diamond truing of vitreous bond SiC wheel are shown in Fig. 7. Werner and Minke [2] showed that G d for polycrystalline and natural diamond truing of vitreous bond aluminum oxide wheels was ranging from 1.2 10 6 to 2.4 10 6. In this study, the G d results, between 0.2 10 6 and 1.3 10 6, is lower but comparable to the data presented by Werner and Minke. The lower G d can be explained by the differences in the grinding wheel. The SiC abrasive used in this study is harder than the aluminum oxide. More importantly, the wheel used in this study has very dense vitreous bond [1], which also reduced the G d. 13 7.2 a 25 13.9 50 27.8 125 69.4 a : test conducted in this test. 4. Experimental results and analysis Werner and Minke [2] defined a dimensionless parameter called the wear ratio, G d. G d = V w V d (2) where V w is the volume of material removed in the grinding wheel and V d is the volume of diamond wear on the tool. Fig. 7. The dimensionless wear parameter G d for the rod diamond blade tool truing of vitreous bond SiC wheel.

592 A.J. Shih, J.L. Akemon / Wear 250 (2001) 587 592 Fig. 8. The wear coefficient G t for the blade diamond tool truing of vitreous bond SiC wheel. Fig. 8 shows the results of G t versus truing traverse speed at four different feeds for two types of blade diamond tools. Unlike G d, G t is not a dimensionless parameter. The unit of G t is mm 2. The use of G t allows the direct comparison of the performance of blade and particle diamond tools. The effects of three variable parameters are discussed in the following sections. 4.1. Effect of truing feed Truing at low feed, in general, has an adverse effect on wear ratio G t. In other words, truing with large feed helps to reduce the wear of the blade diamond tool. This may first seem contrary to intuition because the larger truing feed raises the truing forces, which should increase the wear rate. This could be explained by the brittle fracture of the SiC abrasive and vitreous bond while truing at the high feed. This trend is valid for the range, 1 10 m, of truing feed tested in this study. However, this trend is expected to stop at some truing feed when the diamond tool and the grinding machine can no longer withstand the high truing force at that level. In the range of feeds tested in this study, the results clearly indicate that minimizing the truing feed does not help improve the blade diamond tool wear. 4.2. Effect of truing traverse speed Similar to the trend observed in truing feed, increases in truing traverse speed help to improve the wear ratio G t. The same reason of brittle fracture of the ceramic bond and abrasive is used to explain this trend. 4.3. Effect of type of diamond The results are not conclusive, neither the rod nor the particle diamond tool showed a convincing advantage over the other. 5. Concluding remarks A method to measure the wear of blade diamond tools for precision form truing of vitreous bond grinding wheels was demonstrated. The method was based on the difference in the size of parts ground by a precision grinding machine before and after a set of a repeated truing process. A parameter, G t, was defined to quantify the wear characteristics of two types of blade diamond tools. Experimental results revealed that truing at small feeds and slow traverse speeds does not help to improve the wear rate. The truing force data and the SEM micrographs of the diamond on the worn blade tools in Part II of this paper will further explain the trends observed in this study. Although the rod diamond blade tool does not show definite advantage on the wear rate, this study showed that there are opportunities to design a better rod diamond blade tool. One of the goals is to improve the wear ratio by minimizing the macro-fracture of diamond rods, which will be shown in the SEM micrographs in Part II of this paper. This goal may be possible by increasing the number and reducing the size of the diamond rods in the blade tool. References [1] A.J. Shih, T.M. Yonushonis, US Patent 6,030,277 (2000). [2] F.G. Werner, E. Minke, in: Proceedings of 11th North American Manufacturing Research Conference, Society of Manufacturing Engineers, 1983, pp. 11 17. [3] C.E. Davus, Int. J. Machine Tool Des. Res. 14 (1974) 33 52. [4] T.J. Vickerstaff, Ind. Diamond Rev. (1970), 260 267. [5] K.G. Tkhagapsoev, B.S. Khapachev, Sverkhtverdye Materialy 9 (1987) 30 35. [6] A.J. Shih, Machin. Sci. Technol. 2 (1998) 13 28. [7] A.J. Shih, Int. J. Machine Tool Manufact. 40 (2000) 1755 1774.